Taliban’s Campaign Against IS

Despite a recent decline, the Islamic State (IS), and its South Asian branch IS-K, remains one of the most resilient terrorist organisations on the planet – as recent reports of it planning attacks in Turkey and Europe show. Research carried out in late 2021 to mid-2022 with Taliban and IS members shows that IS-K represented a serious challenge for the Taliban in Afghanistan in this period. While they initially dismissed the threat from IS-K, the Taliban soon developed capabilities to confront it – these capabilities, and IS-K’s responses to them, are the subject of this paper.

The paper outlines five key counter-IS techniques that the Taliban adopted after August 2021: indiscriminate repression; selective repression; choking-off tactics; reconciliation deals; and elite bargaining.

While their initial response was to indulge in indiscriminate repression, the Taliban gradually moved towards an approach focused on selective repression, with the aim of leaving the local communities in areas of IS-K activity relatively untouched. They also considerably improved their intelligence capabilities in this period. By the second half of 2022, the Taliban had succeeded in destroying enough IS-K cells and blocking enough of the group’s funding to drive down its activities and contain the threat. The Taliban also experimented with reconciliation and reintegration, and managed to persuade a few hundred IS-K members in Afghanistan’s Nangarhar province to surrender, contributing decisively to the dismantling of most of IS-K’s organisation there.

However, there were also significant flaws in the Taliban’s approach. This paper finds that their selective approach to tackling IS-K struggled to find firm footing in the absence of a solid system of the rule of law and of external oversight. The Taliban’s leadership appear to be struggling to figure out how to ensure that the lower layers of their security apparatus follow orders to avoid arbitrary violence. The paper further shows how the Taliban have failed to follow through with their initially promising reconciliation and reintegration efforts.

For its part, IS-K showed remarkable organisational resilience in response to the rising tide of the Taliban’s counterterrorism efforts. The group transformed itself into an underground organisation, relinquishing all its bases and moving most of its assets to northern Afghanistan. With this approach, and true to the reputation of its founding organisation, IS, IS-K in Afghanistan managed to survive, even when faced with potentially existential challenges, such as a crackdown on its financial hub in Turkey. IS-K has come increasingly to rely on online activities, including for recruitment.

The Taliban learned faster than most observers expected them to in response to the challenge of IS-K, and scored significant successes. The longer-term prospects of their counter-IS efforts, however, remain dependent on IS-K continuing to struggle financially, because the drivers of mobilisation into its Afghan ranks remain largely unaddressed.

Introduction

The Taliban took power in Afghanistan in August 2021. As practitioners of insurgent warfare, they had to start learning almost overnight ways of doing counterterrorism and counterinsurgency, especially against what emerged as their main challenger, the Islamic State in Khorasan (IS-K). Their early efforts have been characterised as “brutal” and “ineffective”. Others have stated a belief that that the Emirate would not be able to successfully tackle IS-K on its own. As this paper will show, the Taliban initially relied largely on ruthless tactics. However, as shown in a 2023 paper by this author, despite the (very limited) financial means and human resources available, in subsequent months the Taliban’s approach has not been exclusively brutal and at the same time was quite effective, at least in the short term. Indeed, the Taliban, widely seen during their “jihad” (2002–21) as a force of nature, were in reality even then already displaying considerable organisational skills.

This empirical research paper forms part of the EU-funded STRIVE Afghanistan project, and aims to further discuss and analyse how the Taliban applied their organisational capital to countering IS-K. The guiding questions that this paper seeks to answer are: how did the Taliban structure their post-August 2021 counter-IS mix of tactics, how successful were these in fighting IS-K, how did IS-K adapt, and did the Taliban try to achieve long-term stability, seeking non-kinetic approaches and reducing reliance on violence? Since the Taliban do not frame their counter-IS effort with reference to the Western understanding of counterterrorism and counterinsurgency, the author will also avoid referring to such terminologies, and will instead examine their specific tactics. As noted in a rare study of non-Western responses to terrorism, Western theorisations of terrorism and counterterrorism might not be very useful in analysing such efforts by non-Western states and actors.

The discussion focuses on how, after August 2021, the Taliban practised violent repression, both indiscriminately, against people not directly involved in the armed opposition, and selectively, against active insurgents. It also covers how the Taliban have tried to choke off the armed opposition, denying it access to population, supply routes and financial flows. The paper finally looks at whether there may be signs of awareness among the new Taliban elite that their long-term self-interest might be better served by developing reconciliation programmes of some kind, or by reaching some elite bargain.

There are not many large-scale counterterrorism and counterinsurgency efforts that have altogether eschewed all forms of ruthless violence, so analysing a “counter” effort requires some careful qualifications. The first useful distinction here is between selective and indiscriminate violence. A regime that focuses its violence on its enemies can deliver a clear message that those who challenge it will meet a terrible fate, while political quietism (accepting the status quo without resistance) is rewarded. Encouraging quietism while targeting “extremists” (defined as anti-ruling system elements) should therefore be a winning approach, even if utterly violent. The question that follows, then, is why ruling elites should be concerned about achieving anything more than an efficient (selective) repression. This is a pertinent question especially where a violent conflict has already taken off. At that point, some form of repression can no longer be avoided. Following a long-term pattern of indiscriminate violence makes non-violent alternatives hard to buy into for any opponent. However, even choosing selective violence does not necessarily make non-violent alternatives easy to pursue. Different actors within any government will each make their own assessments on where the boundary between violent extremists and quietists may lie, resulting in divisions within a state apparatus and a ruling elite.

Another important distinction is that violent repression may or may not be accompanied by efforts to negotiate local reintegration deals, with the collaboration of local elites. Such deals are often deemed to be a more effective long-term way of stabilising a polity than relying solely on violence, not least because they can potentially create bonds between ruling and local elites, eventually resulting in the latter gaining sufficient leverage with the centre to effectively constrain its use of arbitrary power. Similarly, repression can also be accompanied by elite bargaining, that is, power sharing.

There are also ways of choking off armed opposition with no political concessions and no negotiations, without using extreme violence. Large-scale military deployments, for example, which, in the presence of adequate levels of manpower, can be achieved without reliance on indiscriminate use of firepower, can result in the capture of territory and assertion of control over the population, reducing or denying the ability of the opposition to recruit new members, access sanctuaries, train and transfer supplies. In other words, the aim of such large operations need not be to destroy the enemy, but can be to choke it off. An even better example of choking-off tactics is financial disruption, where violence plays a very small part. These tactics are particularly appealing to ruling elites, but are not necessarily within their reach. It takes an army considerably superior to the opposing forces to monopolise control over territory and population, and it takes a sophisticated intelligence apparatus to block financial flows towards the armed opposition. Moreover, choking-off tactics can be a protracted affair and even an inconclusive one, depending on the skill of the opponents. An armed opposition could continue operating under more adverse conditions even with little or no access to the wider population, and new channels for transferring cash to rebels can always be devised by creative sanctions busters.

This is a reason for ruling elites not to write off political tactics completely. There are other reasons as well for not writing off local reintegration deals and elite bargains. One possible incentive to invest in reconciliation or an elite bargain is the awareness within the ranks of the ruling elite that ruthless repressions, even when efficient in the short term, do not successfully remove the roots of opposition, but instead allow it to resurface generations later, or even sooner, leaving the state vulnerable. Another possible incentive is that repressions can drag on inconclusively and go through critical phases, with the final outcome being uncertain and involving a high cost to the ruling elites. In such contexts, softer alternatives to ruthless repression can gain traction.

This paper is comprised of three chapters. The first examines the state of IS-K and the type of threat it presented to the Taliban as they took power, and how the Taliban assessed that threat. The second chapter discusses in detail how the Taliban sought to meet the IS-K challenge, examining each tactic in turn: indiscriminate repression; selective repression; choking-off tactics; local reconciliation and reintegration; and elite bargaining. The third and final chapter examines IS-K’s response to the Taliban.

To protect sources, neither the names of the interviewees nor their exact roles in their organisations have been disclosed. IS-K interviewees are classified as either “commanders” (leaders of a tactical group of five to 30 men) or “cadres” (district and provincial-level leaders, or managers of support departments such as logistics or finance, among others).

Methodology

With the Taliban–IS-K conflict still under way, any findings of this paper can be only partial and preliminary. There are also clear limitations to the research methodology adopted: research was by necessity limited to oral sources, with limited support from news reports and policy-oriented analysis – which are also often partial – and no access to primary written sources, such as the Emirate’s records, or of course to any internal IS-K documents.

Researching this topic required a number of methodological compromises given that conducting primary research in a Taliban-ruled Afghanistan is extremely difficult. IS-K recruiters and members were, of course, the most difficult to speak to, primarily because they have increasingly been in hiding. As a result, the body of data collected is inevitably incomplete and follow-up on specific themes was often not possible. The analysis contained in the paper inevitably reflects this. However, it should be noted that when reached and given a proper introduction by a third party, such as a relative, friend, colleague or respected individual, even members of IS-K proved quite talkative. This should not be a surprise, as the literature shows that members of violent extremist organisations are typically proud of being members and often brag about their own activities, even when they are supposed to be operating deep underground, as in Europe. The risk faced by this type of research is therefore not one of not obtaining access. There are other risks, however: that interviewees might be affected by a social-desirability bias, resulting in overstating their achievements, capabilities and/or resources; or by reverse causation, leading sources to provide prejudiced information about rival organisations. Mitigation measures are discussed below.

Taliban officials were quite prudent in their answers, but thanks to their internal tensions and differences, Taliban interviewees were also quite often willing to discuss embarrassing details and to acknowledge limitations in their counterterrorism and counterinsurgency efforts. Taliban interviewees were often dismissive of the IS-K threat and overstated the progress made in countering that threat, while IS-K sources did the exact opposite. This was expected, and it was dealt with by interviewing multiple sources within both the Taliban and IS-K, and by spreading the research effort over 20 months, allowing for the time-testing of responses. This was particularly important and useful as it provided validation points for the reliability of the different sources. For example, initialTaliban dismissals of IS-K were proved wrong, as were IS-K’s triumphalist assumptions made in early 2022. The data points provided by sources could only be assessed against one another over time, as in the case of claims about IS-K moving to northern Afghanistan.

While the author takes into account the literature relevant to the topic and the period, this paper relies mainly on empirical data collected through interviews. It is based on a series of 54 interviews, carried out between August 2021 and April 2023. Multiple interviews on both sides of the conflict and with non-aligned individuals, such as elders, clerics, former IS-K members and hawala traders, allowed for greater cross-referencing opportunities. The details are provided in the table below.

▲ Table 1: Breakdown of Interviews. Source: Author generated.

The research methodology was a hybrid of investigative journalistic and ethnographic interviewing. The questionnaires were adapted to each interviewee; there were, in fact, 54 different questionnaires. Questions evolved as knowledge of ongoing trends and developments expanded.

The interviews were commissioned to three Afghan researchers in local languages (Pashto, Dari and Uzbek) and took place mostly in Afghanistan, with some interviews taking place in Pakistan. Two of the researchers were members of the Salafi community, a fact that facilitated access to IS-K sources and reduced risk to researchers to acceptable levels. All of the researchers had a background in journalism and/or research, had participated in previous research projects with a similar typology of interviewees, had been trained to undertake research with a similar methodology, and had contacts or personal/family relations with Taliban and/or IS-K members, which proved crucial in reaching out and gaining access to interviewees.

The risk that respondents might use the interviews to influence external observers or to misrepresent the facts was assumed from the start as a precautionary measure. This risk was mitigated by using different types of interviewees – such as members of either the Taliban or IS-K, elders of local communities where IS-K operates, clerics and traders – who represented contrasting points of view; by interviewing individuals separately and without them being aware of other interviews taking place; and by inserting questions to which the answer was already known, to verify responses. It proved particularly helpful to present interviewees with information gathered from other sources, such as local elders saying that IS-K members were struggling financially, and ask them to comment. Most IS-K sources could not avoid some degree of openness about apparently negative developments concerning IS-K. Public-domain sources, such as media reports and analytical studies, were also used, where available, to check the credibility of interviewees. The researchers chosen did not know one another, to avoid the risk of researcher collusion to manipulate the content of interviews, for example by inventing content to produce whatever they might have believed the project team wanted to hear. This is always a risk when interviews are carried out by field researchers while the project is being managed remotely. The field researchers were also informed that the purpose of the effort was simply to ascertain facts, and that there was no premium placed on specific findings. Finally, the data collected was validated as much as possible via consultations with independent experts and government and international organisations monitoring developments in Afghanistan, who, given the sensitivity of the topic, asked to remain anonymous.

The interviewees were told that their answers would be used in an open-access publication, the type of which was not specified. The interviews were carried out in part face to face and in part over the phone – some interviewees were in locations that were difficult to access. All the interviews have been anonymised and all data that could lead to the identification of interviewees has been removed.

I. The Taliban and IS-K: Sources of Enmity

The conflict between the Taliban and IS-K did not start in 2021. There was tension between IS-K and the Taliban from the moment IS-K was launched in January 2015. By May 2015, the two organisations were at war, competing over territory, but also over the loyalty of hardened jihadists, be they Afghans, Pakistanis, Central Asians or others. The elements most influenced by the global jihadist agenda were those most likely to be attracted by IS-K, even if its Salafist profile discouraged many who would otherwise have been interested. Several hundred members of the Taliban defected to IS-K, contributing much ill feeling. The fighting, mostly concentrated in Kajaki and Zabul (southern Afghanistan), Nangarhar and Kunar (eastern Afghanistan) and Darzab (northwestern Afghanistan), continued throughout the 2015–21 period and led sometimes to atrocities.

In those years, the two rival insurgent organisations had their columns of fighters clashing in a kind of semi-regular warfare. The better-disciplined IS-K had an edge against poorly trained local Taliban militias in 2015–18, but the tables turned in 2019–20, when the Taliban started deploying their crack units against IS-K’s strongholds in eastern Afghanistan. After that and until August 2021, IS-K stayed away from confronting the Taliban head on and sought safety in more remote parts of the east, counting on the fact that the Taliban were still primarily busy fighting the government of the Islamic Republic of Afghanistan.

For some time after the violence between the two organisations started in May 2015, IS-K did not produce much propaganda. It was only in more recent years that IS-K set up a large-scale propaganda campaign against the Taliban, challenging their credentials, both as a jihadist group and their religious credentials, especially what IS-K saw as their lax implementation of Islamic law. Friction between adherents of Salafism, a purist form of Islamic fundamentalism, and Hanafis – Deobandis in particular, but also Sufis – helped to feed the conflict. Although the Deobandis are described as being influenced by Salafism, Salafis see them as practitioners of an impure form of Islam. This is even truer of Sufis. Although IS-K initially downplayed its Salafi–jihadist ideology in the hope of attracting a wider range of supporters, after its appearance in 2015, it gradually took on an increasingly hardline Salafi character. The Taliban, on the other hand, became more and more diverse over time, incorporating, in particular, many members from a Muslim Brotherhood background, while the top leadership remained predominantly Deobandi-influenced, with a strong influence of Sufism as well. While a significant number of Salafis joined the Taliban’s jihad between 2003 and 2015, after 2015, most were attracted to IS-K.

II. Sizing Up the IS-K Challenge in 2021

IS-K’s Manpower

The extent to which IS-K represented a challenge to those in power in Afghanistan, be they the previous regime or the Taliban, has long been a topic of discussion. The UN Analytical Support and Sanctions Monitoring Team, for example, which relies on assessment provided by member states, has provided constantly fluctuating numbers over time. According to IS-K’s own internal sources, IS-K leaders had at their disposal in July 2021 a force of up to 8,000 men. Of these:

• Just over 1,100 were in Pakistan.

• The remaining force was mostly concentrated in eastern Afghanistan (Nangarhar, Kunar and Nuristan), where some 3,700 IS-K members included the bulk of its combat force, some village militias and much of its administrative structure, handling finances and logistics, keeping track of recruitment, making appointments and deciding transfers, planning training and indoctrination, and other tasks. From this area, moving back and forth to and from Pakistan was easy due to the porosity of the nearby border.

• The other important concentration was in the northeast, largely in Badakhshan, with almost 1,200 members in that region. This second concentration included well-trained combat forces and some administrative facilities, but was not very active militarily during this period, and instead sought to keep a low profile in central Badakhshan, chiefly in the Khastak valley.

• Apart from a few hundred IS-K prison escapees, en route to the east, the rest of the force of IS-K (some 1,300–1,400 men) was at this point mainly spread around the south, the southeast, the region surrounding Kabul, the west, and in the main cities, where it operated underground, recruiting or organising terrorist attacks in urban areas. In several provinces, such as Kapisa, Logar, Ghazni, Paktia, Paktika and Khost, IS-K only had a thin layer of some tens of members, tasked with recruitment, intelligence gathering and preparing the ground for expansion in the future.

These figures are largely comparable with those provided by the intelligence services of member states to the UN, which put the membership of IS-K at 4,000 for the latter part of 2021. The figures collated by the UN monitoring committee likely relate to the more visible component of IS-K, that is, full-time fighters based in Afghanistan. As detailed by IS-K sources, of the numbers quoted above, around two-thirds (some 4,600) were fighters based in Afghanistan. A proportion of these were essentially village militias (hence quite invisible to external observers), and a few hundred members of terrorist hit teams.

IS-K sources were claiming mass defections from Taliban ranks in the early months following the fall of Kabul. Such defections would be surprising in light of the morale issues affecting IS-K at that time (see below), and indeed this appears to have been a massively inflated claim. When asked about defections from the Taliban to IS-K after August 2021, IS-K sources had little concrete information to offer and could only cite five lower-level Taliban commanders in Kunar, three in Nangarhar and one in Khost who defected to IS-K. One source in the Emirate’s local apparatus acknowledged that defections from the ranks of the Taliban to IS-K did take place in the early post-takeover months, but had been limited in numbers. The most important defection to be confirmed, at least by local sources, was that of commander Mansoor Hesar with five sub-commanders and 70 fighters in Nangarhar in late August 2021. Another source within the Taliban confirmed only that in the early days post-takeover, two Taliban commanders from Dur Baba and Hisarak defected to IS-K: Mullah Yakub and a’lim Shamsi. Overall, there were few defections (especially when the total manpower of the Taliban is considered), and they added little to IS-K strength and included no high-profile individuals, thus offering little with which the IS-K propaganda machine could work.

IS-K’s Finances

IS-K’s efforts in this period seem to have been marred by financial shortcomings. Sources suggest that the group’s finance operations were badly mismanaged in late 2021 to early 2022. During this period, however, and in line with Taliban allegations, IS-K sources claimed connections with elements of Pakistan’s army and intelligence, translating into logistical help and support for IS-K’s efforts to raise money from “Islamic charities” in Pakistan. It has not, however, been possible to verify these claims.

IS-K Morale

When the Taliban took over, the idea of giving up the fight was reportedly widespread within the ranks of IS-K. Nearly all of the seven former IS-K members interviewed stated that they had been attracted to IS-K to fight “American crusaders”, not the Taliban. This could have contributed to a decline in morale after August 2021 – although respondents might also have wanted to downplay any hatred for the Taliban that they might have harboured. The Taliban also benefited from war weariness in the country, including within the Salafi community. Even elders critical of the Taliban expressed happiness that the fighting had stopped. The defeats that the Taliban inflicted on IS-K in 2019–20 had also left a mark. A further indicator of low morale was the refusal of many detained members of IS-K to rejoin the group after Afghanistan’s prisons were emptied in the chaotic final days of the Islamic Republic. IS-K sources at the time claimed that thousands of escapees from government prisons had rejoined their ranks after the chaos of August 2021. It is clear, however, that, contrary to these claims, many did not rejoin at all, but went into hiding, trying to stay clear of both IS-K and the Taliban (see below on the lack of impact of escapees on IS-K’s strength). It may be added that all former members were reportedly aware that they could contact IS-K via Telegram to rejoin, but many did not take this opportunity. Taliban officials interviewed by the International Crisis Group quantified the escapees who rejoined IS-K in the “hundreds”, rather than in the thousands alleged by some sources.

How the Taliban Assessed IS-K

The Taliban’s initial neglect of the threat represented by IS-K was not due to any form of tolerance. Many senior Taliban viewed IS-K as a proxy organisation, established or manipulated by the security services of the previous regime and/or by those of neighbouring and regional countries, Pakistan in particular, with the intent of splitting the insurgency and undermining the Taliban. The Taliban thought that, with the previous regime gone and the war won, IS-K would be critically weakened by the disappearance of a critical source of support. Moreover, the Taliban’s belief was that IS-K lacked a mass base:

The problem is the Salafi ulema and mullahs, who inoculate the seed of hypocrisy and a very negative view of Hanafism in their Salafi followers … With the normal Salafi villagers, who don’t have any connection with Daesh [IS-K] and with the [Salafi] ulema, [the Taliban’s] relations are very good.

There was also a belief that people had joined IS-K because of the salaries it was able to pay, thanks to generous funding from foreign supporters.

The Taliban leadership, therefore, initially tended to underestimate the threat represented by IS-K. At the same time, while IS-K was not perceived as a strategic threat in August 2021, it was nonetheless considered a resolutely hostile and irreconcilable organisation of “khawarij”, against which the officials of the Emirate were ordered to take “aggressive and serious” action.

III. The Taliban’s Counter-IS Effort

This chapter will discuss the five key counter-IS techniques that the Taliban adopted after August 2021, as outlined in the Introduction: indiscriminate repression; selective repression; choking-off tactics; reconciliation deals; and elite bargaining.

Indiscriminate Repression

The Taliban have in the past argued that indiscriminate revenge-taking and repression on the part of Afghan and US security forces in 2001–04 drove many into the ranks of the insurgency. These views were supported by the elders of insurgency-affected areas. Perhaps because very few local Taliban officials were active with the organisation in those years, they seem oblivious today to the obvious lessons that should have been derived from that experience. Indeed, some Taliban officials have sought to undermine IS-K by trying to crush its supporting networks and milieus. Many Taliban cadres had been fighting IS-K before, and had developed a deep hatred for the organisation, which emerges from virtually all the interviews that the research team carried out. Some also harboured a strong hostility towards the Salafi community, from which they knew the bulk of IS-K’s Afghan members came. Some Taliban equated the Salafi community with IS-K. The fact that the Taliban had experienced serious friction with Salafis since the expansion of their insurgency to the east in 2008–09 helped to strengthen these negative views.

In some cases, indiscriminate repression was a standalone tactic. The best example of this approach in the early wave of post-takeover repression was Kunar’s governor, Haji Usman Turabi, who epitomised the tendency to conflate Salafism and IS-K. Turabi is nowadays acknowledged by members of the Taliban to be “ideologically against Salafism” and to have “killed several Salafi mullahs”. Turabi believed he knew where the main areas of support for IS-K were, and moved to crush local supporting networks and to shut down Salafi madrasas and mosques. All this led to outrage against him, and the Salafi ulema sent a delegation to Kabul to complain.

In other cases, indiscriminate repression was coordinated with other counter-IS tactics. While attempting to undermine IS-K operations in Jalalabad, which was a key centre of IS-K’s campaign of urban terrorism, the Taliban targeted IS-K underground networks and sympathising milieus in Nangarhar. This campaign was initially very violent. A cadre who gained notoriety here for his ruthless approach to IS-K was Dr Bashir, who became head of the Taliban’s intelligence services, the General Directorate of Intelligence (GDI), for Nangarhar province in September 2021, and served in that position throughout 2022. Bashir shut down most of the Salafi madrasas and mosques of Nangarhar. Under Bashir’s leadership, the Taliban in Nangarhar adopted a proactive approach, with large-scale operations and extensive house-by-house searches, detaining many. Many extrajudicial executions of suspects took place under his tenure. The UN Assistance Mission in Afghanistan reported 59 confirmed executions of IS-K suspects, mostly in Nangarhar during October to November 2021. Human Rights Watch indicates that more than 100 suspects were killed between August 2021 and April 2022 in Nangarhar. Salafi community leaders confirmed in February 2022 that in October to November around 100 members of the community were killed in this wave of violence, mostly in Nangarhar. Among them were senior Salafi preachers. Others fled or went into hiding.

It seems clear that Bashir was orchestrating much of the violence, seemingly with the intent of intimidating IS-K support networks and the surrounding milieus – perhaps even the entire Salafi community – into negotiating deals with the Emirate that would guarantee them security in exchange for cutting off relations with IS-K. This approach has similarities with what some of the strongmen of the previous regime had been doing, such as Abdul Raziq in Kandahar, who managed to force local Taliban to negotiate with him after years of relentless and extreme pressure. The Taliban’s reconciliation effort is discussed more fully below.

Selective Repression

The outrage noted above in relation to Haji Usman Turabi’s indiscriminate repression in Kunar led to the Emirate’s authorities deciding to sack him and appoint in his stead Mawlavi Qasim, from Logar, who had served as shadow governor of Kunar during the Taliban’s insurgency (2002–21). Qasim was not popular in Kunar, where the local Taliban base demanded that a local Talib be appointed governor. He appears to have been chosen by Kabul because of his readiness to comply with their request that he avoid unnecessarily antagonising the Salafis, hence transitioning towards more selective repression. The Emirate’s leadership went ahead, even as a very unhappy Turabi threatened to split from the Taliban with his followers.

Turabi’s removal suggests that the leadership in Kabul was seriously concerned about the reaction of the Salafi ulema. However, transitioning towards selective repression was never going to be a smooth path. Even if indiscriminate repression lessened after 2021, much damage had been done, as the repression entrenched the sense in the Salafi community that the new regime posed a critical threat to the community.

Moreover, the new policy of selective repression that followed Turabi’s dismissal was not particularly popular with Taliban officials. Within the Taliban ranks there was denial that indiscriminate abuse had taken place. In the words of a police officer:

The Islamic Emirate always told the normal Salafi villagers [that is, not associated with IS-K] that it doesn’t have any problem with their sect, unless they support the enemy of Afghanistan, the Daesh khawarij … Those Salafi people arrested or killed by the Taliban, they had some kind of connection and relation with the Daesh khawarij.

Even looking forward, doubts persisted that the new policy was appropriate. One GDI officer commented: “I have doubts [about some of the Salafi ulema and mullahs], but we cannot take any kind of action because I don’t have proof … the Taliban leadership in Kabul is trying not to create problems for Salafi ulema and elders in Kunar”.

Some other officials were more explicit in their criticism. As one police officer commented, “The ideologies of Salafi and Daesh are the same, then why they shouldn’t support Daesh?”, implying that the entire Salafi community was a security threat. This officer advocated the closure of all Salafi madrasas and schools and criticised what he viewed as the Emirate’s soft approach, dictated by the fear of driving more Salafis into the arms of IS-K.

Indeed, surrendering IS-K members did warn the Taliban to avoid antagonising the Salafi community, on the grounds that doing so would drive members towards IS-K. Despite this, outside Kunar, Taliban officials continued closing Salafi mosques and madrasas and detaining Salafis, affecting the entire Salafi community. At the end of 2022, Salafi sources alleged that the Taliban had decided to take over Salafi madrasas in southeastern Afghanistan (that is, installing Hanafi principals to run them and replacing many teachers and professors); in universities, teachers accused of being Salafis were dismissed. Taliban and IS-K sources both confirmed these actions. In December 2022, according to Salafi sources, the Taliban took partial control of a madrasa in the Shuhada district of Badakhshan, and in early February 2023, a large-scale Taliban crackdown in Badakhshan led to raids on three local Salafi madrasas and bans on Friday prayers in 10 mosques.

The quantitative and qualitative growth of the Taliban’s GDI was inevitably going to be instrumental in the implementation of the new directives and in making repression more selective. From the start, rather than investing in protecting every possible target from IS-K attacks, the Taliban opted to focus on infiltrating IS-K cells in and around the cities. Given the limited resources available (the entire annual 2022/23 state budget being just above $2.63 billion, or 48% of what it had been in 2020), this appears to have been a sound approach. As a result, a major focus of the Taliban’s effort throughout 2022 was the expansion and consolidation of the GDI’s network of informers throughout the IS-K-affected area. During 2022–23, the Taliban were able to carry out multiple successful raids on IS-K cells, mostly in Kabul, but also in other cities. Dr Bashir was credited with quickly setting up a vast network of informers and spies in the villages and in Jalalabad, which led to the destruction of numerous IS-K cells. The impact appears to have been obvious, as attacks stopped, although other techniques, such as local negotiations and the targeting of supporting networks, were also used (see below). On social media, IS-K repeatedly warned its members about the Taliban infiltrating its ranks, implicitly acknowledging its difficulties.

However, there was some obvious evidence of the GDI’s networks being slow to reach areas where IS-K had not originally been expected to operate. One example is a rocket attack from Hayratan into Uzbekistan on 5 July 2022. This was carried out by three Nangarhari members of IS-K, who were able to hide in a safe house in Mazar-i Sharif for seven months. These outsiders should have attracted the attention of the GDI; the fact that they did not highlights how Taliban intelligence gathering in mid-2022 was still weak in this area.

Another necessary tool for a full transition towards selective repression is the establishment of a functional system of the rule of law. When the Taliban authorities claimed to “have proof” of mosques and madrasas supporting IS-K, including confessions from surrendering IS-K members, such allegations were disputed by Salafi advocates. The Taliban disregarded the advocates’ complaints: “There were some complaints from some Salafi ulema regarding the banning of their madrasas and mosques, but we don’t care”, said one source. In reality, the standards of proof were quite low. A source in the Kunar GDI implicitly acknowledged this: “In Kunar province we have warned Salafi followers that if the Islamic Emirate had a small doubt about any madrasa or mosque spreading propaganda about Daesh, we would close it and will inflict a heavy punishment on the madrasa’s principal or on the mosque’s imam”.

The low standards of proof predictably resulted in the crackdown continuing on and off, even if not as dramatically as before. At least, the excesses of the Nangarhar death squads of October and November 2021 were not repeated on a comparable scale in 2022.

Choking-Off Tactics

In addition to repression, another key approach taken by the Taliban to countering IS-K in recent years has been choking-off tactics. Typical examples of such tactics include cutting off an insurgency’s supply lines, or the financial flows supporting it, or its access to the population. The Taliban should have been familiar with this: one of the major debates between Kabul and Washington in 2006–21 was over the US’ inability or unwillingness to force Pakistan to cut the supply lines of the Taliban. That failure, many argued, made the war unwinnable.

While it would have made sense for the Taliban to destroy IS-K’s bases in the far east of Afghanistan in order to disrupt the group’s ability to maintain its influence in eastern Afghanistan, they had limited manpower available as they were taking over the Afghan state in the summer of 2021, with just some 70,000 men in their mobile units as of September 2021. The Taliban’s Emirate had to concentrate thousands of its best troops in Panjshir from early September 2021, where it faced the resistance of local militias and remnants of the previous regime’s armed forces, gathered into the first new armed opposition group to rise after the regime change, the National Resistance Forces. Thousands more troops were busy securing the cities and sealing the border with Tajikistan. The scarcity of manpower in this period is highlighted by the fact that in the months following the takeover, there was only a very thin layer of Taliban armed forces present in most rural areas. In the average district, the Taliban were only able to deploy 20 to 30 men, who guarded district-centre facilities and carried out occasional patrols, riding motorbikes on the roads. They were rarely seen in the villages. While a process did begin of Taliban supporters, reserves, sympathisers and relatives of Taliban members joining the Emirate’s armed forces, it took several months to absorb these untrained or poorly trained individuals into the forces. Moreover, plans for the security sector were initially quite modest, as the Emirate’s leadership decided to keep the size of its armed forces relatively small, for several reasons:

• The easy victory obtained by the Taliban in Panjshir in September.

• The fact that IS-K was viewed as a marginal actor due to its low profile (see “How the Taliban Assessed IS-K”, above).

• The positive attitude shown to the new regime by all neighbouring countries, except for Tajikistan (which was hosting the National Resistance Forces), was making it hard for armed opposition groups to find a safe haven and external support.

• The limited fiscal base of the Emirate.

Indeed, Taliban sources circulated the news that the new army would be small, with as few as 40,000 men in combat units and another 20,000 in support and administrative roles. The police force was planned to be 40,000–60,000 men, of whom some 5,000 would be in a special force called Badri 313. These plans soon changed, however, and by January 2022 the Emirate had upgraded its plans for the army and police, overseeing a gradual expansion of the army towards a target of at least 150,000 men. It seems likely that the resumption of IS-K activities in the cities and in the east contributed significantly to this decision.

The Taliban therefore delayed launching any large operation in the east. They seem to have understood that large military sweeps without the ability to hold territory afterwards are pointless, if not counterproductive – possibly as a result of having observed the failure of such tactics when used against themselves before August 2021. By March 2022, the Taliban were finally able to launch their first relatively large operation in Kunar, with the intent of forcing IS-K to fight for its bases. Initially, they seem to have thought that by threatening the few fixed bases IS-K had in the far east, they would force IS-K to stand and fight, and inflict major losses. According to a local Taliban source, before August 2021, IS-K had access to “every district of Kunar” and had “very active military bases and training centres”. But the insurgents avoided contact, leaving their bases behind and pulling deeper and deeper into the upper valleys. A Salafi a’lim (religious scholar) offered a similar assessment for Dangam district, saying that IS-K had controlled about 30% of the territory before the Taliban takeover, but that most IS-K members moved out after August 2021. The GDI expected to need another military operation, even deeper into the valleys, to “finish IS-K off”. By April 2022, however, the Taliban realised that IS-K had given up its last vestiges of territorial control in Kunar without a fight.

Whether or not this was initially part of their plans, the Taliban considered that they had achieved an important objective: although IS-K tactics made it impossible for the Taliban to eliminate the group, asserting control over territory and population would still allow them to choke off IS-K. A Taliban cadre in Kunar said in April 2022 that IS-K’s opportunities to approach potential recruits had been greatly reduced, as it had been forced to go underground and to downscale operations.

The Taliban’s pervasive presence on the ground also allowed the GDI to improve its mapping of IS-K’s presence countrywide. By March 2023, for example, the Taliban claimed to fully know where IS-K cells were operating in Kunar. This choking-off tactic therefore also contributed to enabling more selective repression.

The other main choking-off tactic used by the Taliban against IS-K was financial disruption. Hawala traders were saying in late 2021 and early 2022 that Taliban authorities (the GDI, but also the National Bank) were increasing pressure on them. At that time, the Taliban had not yet worked out how to effectively block hawala traders from transferring money for IS-K (or any other hostile actor), and so relied on intimidation and implementing existing rules for registering transactions – woefully ignored under the previous regime – to achieve impact. Visits from Taliban patrols served as reminders of the danger of cooperating with IS-K. While these tactics could not completely stop the flow of cash for IS-K from Turkey (where the main financial hub of IS-K was located), they do not seem to have been pointless. IS-K sources reported that by September 2022, IS-K could only rely on a very limited number of hawala traders and a few smugglers who were taking cash for IS-K from Pakistan into Afghanistan. Later in the year, financial transfers were complicated further by a Turkish government crackdown on IS-K networks in Turkey. It is not clear whether the Turkish crackdown was the result of intelligence provided by the GDI, or of the Emirate’s “diplomatic” engagement. In any case, as an IS-K source acknowledged, the group’s expansion into the north was insufficiently funded as a result.

These efforts appear to have had some impact. One IS-K source claimed in May 2022 that earlier financial flow problems had been fixed, but there was evidence to the contrary. Salaries paid to frontline fighters, at $235 per month in 2022, were lower than in 2015–16, when they were reportedly as high as $600. Although the central leadership of IS continued to promise massive funding increases for the future, in 2022, according to one of IS-K’s financial cadres, it cut the IS-K budget to its lowest level ever.

The Taliban’s Reconciliation Deals

As noted above, Dr Bashir was not simply interested in wreaking havoc in IS-K-supporting networks and milieus. Having gained a position of strength through his crackdown, Bashir moved forward with local negotiations with community elders to undermine the rival organisation. The Taliban had themselves been subject to reconciliation efforts to co-opt some of their ranks when they were fighting their “jihad”, although it is not clear what they made of these efforts, which were in any case poorly implemented by the Afghan government of the day. Bashir is now seen by Taliban officials as having been a “very active chief for Nangarhar GDI department” and as having had a “very good connection with villagers and elders in every village and district of Nangarhar province”.

The Taliban were probably aware of the role played by Salafi elders in the recruitment of IS-K members, or perhaps presumed such a role, based on their own experience as insurgents. Several surrendered IS-K members acknowledged that many Salafi elders in Nangarhar had previously encouraged villagers to join IS-K. IS-K teams had regular meetings with elders, encouraging them to mobilise villagers. There was reportedly a high level of pressure on individual members of IS-K to invite friends, relatives and neighbours into the group. It was standard practice for Salafi village elders supporting IS-K to be trusted to introduce new members without the standard additional vetting. “Joining Daesh at that time was very easy; it only needed one telephone request”. Individual recruits, on the other hand, were still scrutinised much more seriously, according to a former IS-K member who was recruited via social media.

Dr Bashir relied on an initially small number of Salafi elders willing to cooperate, and on several Hanafi elders who had connections with some IS-K members or lived in areas affected by the IS-K presence. Former IS-K sources confirm the role of the elders in negotiating their surrender. In the words of one, “When we decided to surrender to the Taliban’s Islamic Emirate, again we used the local elders to negotiate and mediate our surrendering with Dr Bashir”. The GDI arranged for the surrendering IS-K members and their community elders to guarantee under oath that they would not rejoin IS-K or in any way oppose the Emirate. The elders agreed to take responsibility and inform the Emirate’s authorities of any violations.

On the basis of Dr Bashir’s exploratory efforts in 2021, the GDI and other components of the Taliban’s security apparatus established communication with community elders. The village elders were tasked by the Taliban GDI with negotiating the surrender of any Salafi elder with whom they came into contact. The Taliban identified useful contacts among the elders, and the district governor or the chief of police regularly visited them, as often as weekly or fortnightly.

The official claim is that in 2021–22 some 500 IS-K members (commanders, fighters, recruiters, support elements and sympathisers) from Nangarhar, Kunar and Laghman surrendered as a result of Bashir’s combination of ruthless repression and negotiations with the community elders. This figure is likely to be somewhat inflated. One of the surrendering IS-K member noted that “there were lots of people among those 70 who surrendered who were not Daesh members; I didn’t recognise many of them”. A source in the Taliban’s provincial administration acknowledged that some Salafi elders, anxious to please the new regime, convinced some members of the community to pose as IS-K members and “surrender”. This was discovered later by the GDI but, overall, the elders-focused effort was still rated highly successful. A police source estimated that 60% of those surrendering were IS-K members from eastern Afghanistan and 40% were civilian supporters. Even a source hostile to the Taliban supported a positive assessment of the campaign, acknowledging that in a single village in Sorkhrod, three IS-K members surrendered to the Taliban. Various ex-IS-K interviewees confirmed having surrendered as part of large groups of IS-K members.

The majority surrendered because of agreements between the GDI and community elders, but some surrendered directly to the GDI, after Bashir managed to reach out to them in the districts and convince them that surrendering was the best option for them. Bashir’s argument to these IS-K members was that it was not in the Salafi community’s interest to have another war, which would be fought ruthlessly by the Taliban, including in their villages.

With a much reduced IS-K ability to threaten waverers, due to the group’s weakness on the ground, the path was clear for the Taliban to expand their tactics of negotiating deals with community elders to Kunar province. Indeed, to some extent during 2022 the stream of surrendering IS-K members, which had started in Nangarhar in autumn 2021, spread to Kunar. Here too, the Taliban sought the cooperation of the community elders to convince IS-K members to lay down arms. Some Salafi ulema were also involved. Although the surrenders were fewer than in the neighbouring province, the “tens of Daesh members” who surrendered to the Taliban as a result of the mediation of the elders represented a warning to IS-K. The formula adopted was the same as in Nangarhar, with surrendering members taking an oath never to rejoin IS-K and the elders guaranteeing for them. As in Nangarhar, some IS-K members in Kunar reached out directly to the GDI to negotiate their surrender.

At the same time, the Taliban continued their local negotiations with elders in Nangarhar. The flow of surrenders therefore continued in 2022. The last group to surrender in 2022 was composed of some 70 members from Nangarhar, who defected in the autumn. As of January 2023, the Taliban believed that 90% of the IS-K structure in Nangarhar had been wiped out; the Taliban were aware of the existence of some IS-K cells, but deemed them too weak to launch attacks. It is difficult to say whether the Taliban’s estimate was correct, but undoubtedly IS-K had taken a big hit in Nangarhar.

Those who laid down weapons sometimes reported being treated decently by the Taliban; others reported not being treated very well, with Taliban and pro-Taliban villagers looking down on them. Still, they appreciated that they could live with their families, even if most of them had had to relocate to avoid IS-K retaliation. There were complaints about being required to report to the police station every week or two, and not being allowed to move around without permission. Surrendered IS-K members also complained that the Taliban were not implementing their side of the deal – specifically, giving financial support to those who had surrendered. One of those interviewed noted that this would make it hard for the Taliban to convince more to surrender. Another complaint was that those who stayed in the districts did not feel safe from IS-K.

The fact that madrasas and some mosques were still closed also upset the reconciled IS-K members, in part because the surrender agreements included a clause about reopening them. Reportedly, the surrendering IS-K members had been promised government jobs, the freedom to live anywhere in the country and the receipt of cash payments for six months. In practice, no cash was paid (although some food and some benefits in kind such as blankets were provided), and the surrendering men were only allowed to choose to live in their own community or in the district centre. Some surrendered IS-K members hinted that the reason why surrenders have slowed down was to be found in the violation or non-implementation of these agreements.

Elite Bargaining with the Salafi Ulema

In 2020–21, the Taliban did not show much faith in the opportunities offered by intra-Afghan talks, nor were their counterparts in Kabul able to pursue those talks with any degree of effectiveness. Instead, the Taliban sought to co-opt local and regional elites associated with the government of the Islamic Republic. It is probably in a similar spirit and informed by this experience that the Taliban approached the prospect of negotiations for resolving the conflict with IS-K. The Taliban were well aware of the links between IS-K and much of the Salafi clergy. Support from Salafi communities in the east and northeast had proved essential for IS-K to be able to put down roots there. Many Salafi preachers were recruiting for IS-K in this period, as sources within the community admit, and Salafi madrasas and schools in Kabul were sending numerous recruits to IS-K. Much of the Salafi youth joined during this phase. For the Taliban, driving a wedge between IS-K and the Salafi community, from which the former draws most of its support base, must have seemed an attractive opportunity.

A group of Salafi ulema had already sought an understanding with the Taliban in 2020, as IS-K was losing ground quite fast in the east. A delegation of senior Salafi ulema, led by one of the most senior figures, Sheikh Abdul Aziz, met the Taliban’s emir, Haibatullah Akhundzada, and other senior Taliban in 2020, offering support to the Taliban in exchange for the cessation of violence and reprisals against civilians. The Emirate’s authorities again welcomed delegations of Salafi ulema in Kabul in 2021, reconfirming the agreement with the Salafi ulema and reissuing orders that the Salafis should not be targeted. After that, attacks and harassment of the Salafis reduced, even if some Taliban commanders continued behaving with hostility towards Salafis.

However, the terms of the agreement were that the Taliban would not allow the Salafi preachers to proselytise, and the madrasas that had been shut on grounds that they had been recruiting for IS-K remained closed. Only the mosques were reopened. Moreover, some senior clerics, accused of links to IS-K, remained in prison: Sheikh Bilal Irfan; Sheikh Qari Muzamil; Sheikh Sardar Wali; Sheikh Jawid; and Delawar Mansur. The Salafi ulema interpreted the closure of the madrasas as temporary and expected that after some time the community could return to its quietist stance, which had in the past (before 2015) been the predominant position among Afghan Salafis.

Nevertheless, even after the second agreement in 2021 many “hot-headed” young members of the community stayed with IS-K. One of the Salafi ulema pledging allegiance to the Emirate admitted in a private interview that the Salafi clerics remain opposed to Hanafi Islam, but that they did not think IS-K stood a chance against the Taliban, and that it was not in the interest of the community to fight. These clerics, however, did not have control over the youth who were still with IS-K.

On the other side, among the Taliban and the Hanafi ulema, there were voices of moderation, especially among the ulema, who were even willing to tolerate Salafi proselytising – generously funded from Saudi Arabia and Pakistan – on the grounds that otherwise the Salafis would continue being driven towards IS-K. An imam in Jalalabad expressed what might be defined as the midway solution preferred by the Taliban’s leadership, as discussed above: avoid identifying all Salafis as linked to IS-K; leave the Salafis alone; but ban them from proselytising. His words reflected angst about the seemingly unstoppable spread of Salafism: “I am living among Salafi scholars and followers; they are becoming bigger and bigger every day, they have very good financial sources in Saudi Arabia and several other Arab countries … to expand their activities”.

But the 2021 agreement was also opposed by many among the Taliban and the Hanafi ulema. There are many hardliners. Former Kunar governor Turabi embodied the hardline stance: repression without local reconciliation efforts. Although this approach was not effective and was opposed in Kabul, within the GDI’s ranks, Turabi still had supporters in early 2023, who argued for a crackdown on supporting networks and milieus on the grounds that the safe haven they offered was essential for IS-K operations.

A common view among Hanafi ulema is that while there are quietist Salafis in Afghanistan who have not embraced the militant Salafism of IS-K, the popularity of IS-K among Salafis is not only due to a defensive reaction on the part of the community. They believe that jihadist Salafism has been spreading through the community. Because of this, many Hanafi ulema have been sceptical about the decision of a number of high-profile Salafi clerics to seek an understanding with the Taliban, believing it to be only a tactical decision to buy time.

As a result of polarised views within the Taliban and among the Hanafi ulema, the policies of the Emirate concerning the Salafis have continued to fluctuate and vary from province to province, as discussed above. As a result, relations with the Salafi community have remained tense. Kunar received special treatment, with the Taliban’s leadership making clear that especially in Kunar, the GDI should only act against Salafi madrasas and mosques in the presence of solid evidence. The new policy of “working hard to give respect and value to our Salafi brothers and trying our best to finish the dispute between Taliban and Salafi” was introduced after Turabi’s dismissal, according to a source in the provincial administration. The decision was made at the top: “Taliban local leaderships have been told by our leaders in Kabul to keep a good behaviour with Salafi members in Kunar”. There was an at least partial acknowledgement that “one of the reasons why Daesh in Afghanistan became active and somewhat powerful is that some Taliban carried out aggressive acts against the Salafis in Kunar and Nangarhar”. Former IS-K members confirmed that negotiations with Salafi elders and the ulema led to the reopening in 2022 of all mosques and of the Salafi madrasa, except two, which stayed closed due to their connection to IS-K.

Despite this “special treatment”, a Salafi a’lim estimated in April 2023 that the community in Kunar was split between those who have functional relations with the Taliban and those who are hostile. One Salafi elder estimated that in his district of Dangam, 30% of the Salafi community was on friendly terms with the Taliban and the remaining 70% had tensions. It did not help that the Salafis remained marginalised in Kunar even in early 2023, as all the provincial officials were Hanafi, with only a few rank-and-file Taliban from the Salafi community. The Taliban have regular meetings with the district shura (council) and occasional meetings with the village shuras, but no Salafis were included in the district shura or in at least some of the village shuras. Hence, a Salafi elder complained that “the Taliban don’t want to hear too many complaints from the Salafis, nor their views”. Clearly, while attempting to defuse tension, the Taliban seemed to have no intention of moving towards an elite bargain.

Even Taliban sources acknowledge that friction between Salafis and Hanafis has persisted. For example, throughout 2022–23, the Taliban were insisting that all imams wish a long life to the Taliban’s amir (or “head of state”), Mullah Haibatullah Akhundzada, during Friday prayers; the Salafi ulema in Kunar refused to comply. This refusal did not lead to a new crackdown, but it shows that the Salafi ulema were not entirely committed to supporting the Emirate, despite their pledge. The Taliban had offered them a safety guarantee as subjects of the Emirate, but it appeared that the Salafis wanted an elite bargain, that is, at least a share of power and influence. As a result, the Taliban’s engagement with the Salafi ulema went cold towards the end of 2022. After two or three meetings during 2021–22, meetings stopped, and Taliban officials took the view that the Salafi ulema were not willing to fully implement their part of the deal and that several of them were still supporting IS-K.

There appears to have been no talk at any stage of incorporating significant numbers of Salafi clerics into the ulema councils at the provincial and national levels, which would have been a major step towards an elite bargain with Salafi elites.

IV. IS-K’s Response to the Taliban’s Tactics

While the Taliban’s efforts posed major challenges to IS-K, not all the techniques discussed above were threatening or, indeed, were perceived as such. IS-K does not appear to have been concerned about indiscriminate repression against its supporting milieus, and its only apparent reaction was intensifying efforts to present itself as the defender of the Salafi community. Its focus was instead on responding to the Taliban’s choking-off effort, especially their campaign to take full control of territory and population.

The Response to Choking-Off Tactics

Even if the Taliban were not, immediately after their takeover, in a position to organise a major military campaign in the far east of Afghanistan (Kunar and Nuristan), IS-K clearly understood the potential threat this would represent. By the time the Taliban took over in August 2021, IS-K had long opted out of a direct confrontation with them, after it had emerged in 2019–20 that its forces could not stand up to the Taliban on an open battlefield. This perception of a major threat from a Taliban assault on IS-K bases in the far east only increased after August 2021, given that the Taliban were at that point no longer busy fighting the forces of the previous regime. IS-K soon relinquished the residual territorial control it still had (see the discussion of choking-off tactics above). The group appears to have hoped to delay the expected Taliban onslaught in the east, or to make it unsustainable by waging a guerrilla war against the Taliban forces deployed there, forcing them to divert forces – while at the same time mitigating the impact of choke-off tactics by reducing the number of non-local members (who were harder to hide and more difficult to support) and creating an extensive underground network.

Delay and Diversion

While seeking to retain control over parts of Kunar and Nuristan, IS-K largely switched to asymmetric tactics, such as intensified urban terrorism, hit-and- run raids, ambushes and mines. These efforts produced few results initially, and IS-K’s leaders (the leader of IS-K and the military council) had to keep thinking of new strategies. A plan for sending cells to cities where IS-K was not yet active, such as Kandahar, Herat and Mazar-i Sharif, was hatched in spring 2021 – that is, before the Taliban took power – although it was not fully implemented until August 2021.

Essentially, the IS-K leadership decided to keep the Taliban busy by going on the offensive in the cities, calculating that by risking a few tens of cells it could force the Taliban to commit tens of thousands to guarding the cities. The campaign started somewhat slowly, due to the limited capabilities of existing IS-K underground networks in Kabul and Jalalabad.

During the last five months of 2021, IS-K was able to increase the number of its large terrorist attacks in Kabul to five, from two in the first half of 2021. Urban guerrilla actions also continued in Jalalabad after a short lull, opening up with a series of six bomb attacks in September, followed by some months of urban guerrilla warfare against members of the Taliban. Taliban sources described the situation in Jalalabad at that time as “daily IS-K attacks”.

At the same time, during the chaotic power transition of summer 2021, IS-K was able to transfer multiple cells to the cities, which reinforced its presence in Kabul and Jalalabad but also allowed it to expand its terrorist campaign to cities previously unaffected by this campaign. Cells were thus established in Kandahar, Herat, Mazar-i Sharif, Charikar, Kunduz, Faizabad and Gulbahar. Among the cells were recruiting teams which targeted, in particular, university campuses. As a result, while IS-K was able to intensify its campaign of terrorist attacks in the cities, it was also hoping that the new urban underground structure would become self-sustainable. An IS-K source acknowledged that the group exploited the chaotic period of the Taliban’s takeover to send more of its cells into the cities. He explained that “because different groups of Taliban entered Jalalabad city and other cities of Afghanistan from the mountains and the districts, it was very difficult for the Taliban … to distinguish between Daesh and Taliban members there”.

An IS-K source estimated in early 2022 that the Kabul city contingent, following years of decline, had climbed back up to 300 members, in two separate structures – one aimed at preparing and carrying out attacks, and the other at recruiting and propaganda operations. There seemed to be a real opportunity for catching the new regime off guard, with the Taliban still surprised to find themselves in power and dealing with multiple crises in their efforts to keep the Afghan state afloat. While the Taliban were known to be more than a match for IS-K in a conventional fight, IS-K hoped that the Taliban’s lack of experience in counterterrorism would allow several hundred terrorists to cause havoc in the cities, as even Taliban officials confirmed to the International Crisis Group that this was the case.

Aside from its intensity, in terms of target selection the campaign of terrorist attacks in Kabul was a continuation of IS-K’s earlier campaign against the previous government. The targets of the new phase of the campaign were also religious minorities, such as the Sikhs and, most of all, the Shia community. Aside from forcing the Taliban to divert forces away from the east, the primary intent seems to have been to create chaos in the cities, turning the sizeable Shia community against the Taliban (for their failure to protect it) and exposing the incompetence of the new regime, especially in urban security. In spring 2022, the High Council of IS-K decided, in the context of some fine-tuning of its strategic plan, to further reinforce the focus on terrorism in the main cities, targeting the Shia community via a wide selection of very soft targets, such as schools and mosques. Protecting so many potential targets would have required the Taliban to commit significant human resources, to the detriment of the wider counter-IS effort.

Operationally, IS-K’s campaign in 2022 produced some visible results. According to a respondent, IS-K’s “research and inquiry” department, which undertakes analysis for the leadership, produced in June 2022 an internal report indicating that in the spring of 2022, IS-K had achieved the highest number of “highlight” (that is, headline-making) attacks and military activities in three years. Impartial data collection shows that the pace of bomb attacks peaked above 10 per month in April–July 2022, but started declining in the latter part of that year, to between three and six per month (see Figure 1). This might have been due to increasingly effective Taliban counterterrorism. However, it is also likely that relocation from the far east had largely been completed, and that IS-K downscaled terrorist attacks in Kabul to a more sustainable level.

▲ Figure 1: IS-K Activity and Taliban Counterterrorism Operations, 2022–23. Source: Afghan Witness, “Taliban Continue Raids Against ISKP in May, Claim Killing of Deputy Governor in Kabul”, 1 June 2023. In the figure, “Arrests” and “Clashes/Raid” refer to Taliban operations against IS-K. Reproduced with permission.

Mitigation

To lessen the need for supplies inside Afghanistan and also being increasingly unable to protect non-Afghan members, in late 2021 and early 2022, IS-K moved more of its Pakistani members across the border. Taliban sources too noted the disappearance of not only Pakistanis but also Central Asians, Chechens and other non-Afghans from the east, and assumed they too had crossed the border.

The process of evacuating the bases in the east took eight months; even for some time after this a substantial number of IS-K members, especially leadership and administrative cadres, were hiding in caves and other secret locations, while their relocation was being arranged. The permanent bases were replaced during 2022 by an underground infrastructure, not only in Kunar but also in parts of Nangarhar, with secret cells established in Achin, Naziyan, Lal Pur, Pachir wa Agam, Bati Kot, Mohmand Dara and Jalalabad city. Even as the Taliban kept destroying its cells in Jalalabad, IS-K was able to maintain a presence there. Local elders confirmed the disappearance of obvious signs of IS-K presence, but believed that the group maintained secret cells. In January 2023, a source in the Taliban’s administration stated that IS-K’s presence in Nangarhar consisted of some IS-K cells in Jalalabad and one to two cells each in some districts, such as Achin and Naziyan. As of March 2023, the police estimated that there were 16 IS-K cells in Jalalabad, based on the confessions of detainees, but the cells operated independently and tracking them down was difficult.

Parallel to the move underground, IS-K also sought to adopt a mobile infrastructure to support the small, dispersed cells, a process that continued throughout 2022. A year after the spring 2022 strategic shift was decided, one IS-K source described as an accomplished fact a new, leaner and more mobile infrastructure that had replaced the old fixed bases:

Daesh has training centres and lots of secret cells and secret military bases in Kunar province, but they are changing their locations all the time. Daesh is on the move – its training centre, military bases [and] secret cells are all moving and changing every three or four months. When a member of Daesh is arrested by the Taliban or surrenders, Daesh immediately finds out where these guys were trained, which posts or secret cells they were assigned to, then it changes the locations.

Taliban sources confirmed that IS-K was moving people to the northeast and north and even claimed that the collapse of IS-K activities in Nangarhar was in part due to IS-K moving out.

While IS-K implemented these mitigating actions quickly, it remains the case that they were not enough to prevent the group’s operations from being constrained. IS-K’s messaging to its members did not mention the coming downgrade of the east, for good reasons. It appears to have been a difficult decision to take, given that a large majority of the group’s Afghan members were from the east and had families there. As of early 2022, IS-K sources were still adamant that they would soon go on the offensive, that their bases in the east were safe and that they had enough manpower to defeat the Taliban in the east. The rationale for having IS-K’s main bases in eastern Afghanistan (Nangarhar, Kunar and Nuristan) was still being promulgated by IS-K sources at least until mid-2022: “there are many Salafi people and madrasas in these provinces and most of the followers of Salafism are supporting IS-K”. It took until 2023 for IS-K sources to begin showing awareness and acceptance of the fact that IS-K had given up any ambition to hold territory, at least in the short and medium term.

The constraints that the transition placed on IS-K’s operations are evident when we look at its guerrilla operations in the east. While the transition was ongoing, IS-K, remarkably, sought to keep waging a guerrilla war in eastern Afghanistan. The guerrilla campaign was always limited in scope, affecting only the provinces of Kunar and, to a lesser extent, Nangarhar. Guerrilla activities intensified from late summer 2021, especially in Ghaziabad, Naray and Shegal. Though these mostly consisted of small hit-and-run attacks on Taliban posts and small ambushes, they were beginning to annoy the Taliban. In spring 2022, the High Council of IS-K, while deciding to intensify the terrorist campaign in the cities, also confirmed the decision to continuing the guerrilla war against the Taliban, where possible. However, the new structure left behind in the east proved unable or unwilling to support a steady insurgency there. IS-K guerrilla attacks in Nangarhar remained especially rare. One of the last few recorded attacks was in February 2022, an ambush in Achin which killed two members of the Taliban.

In Kunar, the picture was similar. In one of the worst incidents, a convoy was ambushed in Shegal and “several Taliban fighters were martyred”. In Dangam in Kunar, some lingering IS-K presence continued in the forested area, without much military activity. Those remaining were local members, reportedly being kept in reserve and perhaps supporting the planning of attacks elsewhere. Most IS-K members had reportedly moved to northeastern and northern Afghanistan (see below). This is likely to have affected the pace of guerrilla operations in the east, not only because of lower numbers, but also because to local members the option of lying low and hiding was more likely to seem viable than it would to their foreign and out-of-area comrades. As the presence of non-local fighters dried out, the level of guerrilla activity declined further. An independent assessment found that IS-K was able to sustain the number of guerrilla attacks at between five and 10 per month during the first half of 2022. The numbers, however, collapsed to between two and five in the second half of the year (see Figure 1, where guerrilla attacks are listed under the category “Gun”).

IS-K also tried to adapt in response to the Taliban’s financial disruption operations. Confronted with the news that IS-K networks in Turkey had taken a major hit, IS-K sources indicated that the organisation coped successfully, reactivating its old financial hub in the UAE, where the abundance of Afghan hawala traders would make it easier to find complicit ones. The source had to acknowledge that there was a bottleneck at the receiving end, in Afghanistan, as hawala traders were wary of getting caught. He tried hard to present an optimistic picture, noting that other ways of transferring money, through complicit businesses based in Turkey and through flights between Istanbul and Kabul, with the help of some personnel at Kabul’s airport, were being tested. One of his colleagues also suggested that the financial strangulation of IS-K was lessening as of December 2022–January 2023.

The Response to the Reconciliation and Reintegration Deals

The other main concern for IS-K appears to have been about countering the Taliban’s local reconciliation and reintegration efforts, which had the support of some Salafi elders in the villages (see discussion above). The group appears to have seen this as the biggest medium-term threat. IS-K started in 2021–22 to bring pressure on the elders not to facilitate negotiations between IS-K members and the Taliban. One surrendering member heard from villagers that “Daesh is trying a lot to undermine this process. Several elders who were secretly facilitating the negotiations and connecting IS-K members with the Taliban for their surrender have been threatened”.

Others who surrendered confirmed the same, adding that threats consisted of death threats and threats to burn down the homes of anybody making deals. One of the surrendered members claimed he and two fellow former comrades in arms received threats from IS-K; the group, he said, threatened to “set fire to my house and throw me into the blaze”. Two elders of his village, who had helped the Taliban, he said, were also threatened, and as a result stopped being involved in negotiating surrenders. One even reported that nine surrendered IS-K members ended up rejoining IS-K in Nangarhar, although it is not clear whether this was because of the threats or because of the poor Taliban implementation of the deals. IS-K also increased counter-intelligence efforts among its own ranks. These countermeasures were deemed to be effective by a number of former IS-K members, who believed that surrenders were diminishing or even ceasing. This suggests that IS-K feared the reconciliation/reintegration plans much more than it did indiscriminate repression.

The Response to the Taliban’s Tentative Elite Bargaining

Because of the lack of Taliban success in negotiating with the Salafi ulema, IS-K may not have considered a response to their negotiations with the Salafi ulema a priority – although it is likely that it brought pressure to bear on the Salafi ulema to stay away from the Taliban. IS-K’s short campaign of attacks on pro-Taliban clerics in the summer of 2022 might also have been intended to provoke Taliban retaliation against Salafi clerics and spoil the Taliban’s discussions with them. The killing of Rahman Ansari in Herat in September 2022 might have been a warning as well, as Ansari was a Salafi preacher who had pledged loyalty to the Taliban. IS-K did not claim the killing. The campaign was abandoned in autumn, probably as it was becoming clear that IS-K did not need to be concerned about Taliban negotiations with the Salafi ulema.

IS-K Counterattacks

While IS-K sought to counter Taliban tactics or at least to limit the damage, its leadership also decided to try re-seizing the long-lost initiative by striking the Taliban where it felt they were more vulnerable. The urban terrorism campaign, discussed above, was more of a diversion than a counter-offensive. Instead, IS-K appears to have placed its hopes for turning around the situation in its expansion in the north. Plans to expand recruitment in the north started in mid-2020 (after an earlier aborted effort in 2017–18). Small numbers of Afghan Pashtuns and even Pakistanis were also sent north. After 2021, these efforts were strengthened, and even moving the IS-K headquarters there in the future was considered.

In mid-2022, the IS-K leadership was reportedly still in Kunar, but the new phase of the transfer to the north had been initiated a few months earlier. The movement of people and assets to the north and northeast continued, as both a Taliban police officer and a local elder confirmed. IS-K sources talked up the migration with the claim that it was about taking jihad to Central Asia. IS-K sources spoke about training centres being established in Badakhshan, Kunduz and Jawzjan, with plans to open one in Balkh. As IS-K also dramatically expanded its social media activities, it began releasing significant quantities of propaganda, such as statements and pamphlets in Uzbekistani, Tajikistani and Uyghur, in order to support its claims of imminent expansion into Central Asia.

IS-K seems to have had expectations of rapid expansion into Faryab and the northwest in spring 2022, exploiting intra-Taliban friction. More generally, it is clear that one of the main reasons for the shift in focus northwards was the hope for major defections from the ranks of the Taliban there. That did not happen on any significant scale. When asked for details, an IS-K source could only provide modest defection figures for the entire August 2021 to mid-2022 period: “a few commanders in the north”, with some more in talks as of mid-2022.

Another aspect of IS-K’s “counter-offensive” was to make up for the group’s limited achievements with media-focused symbolic attacks, such as rocket attacks from Afghan territory on Uzbekistan and Tajikistan, which caused no damage but won high-profile exposure in the media. An important part of IS-K’s strategy was integrating its military and propaganda campaigns. Graphic details of the terrorist campaign were used by IS-K social media propaganda to project an image of strength and power that was out of all proportion with the reality. Overall, the leadership of IS-K succeeded fairly well in hiding the extent of its difficulties. The regional and world media, as well as policymakers, continued to portray it as a highly threatening organisation, even though its military achievements were almost negligible.

Although it is difficult to measure how IS-K members and sympathisers reacted to this propaganda, it is clear that one of the intents was to shore up the morale of increasingly dispersed members and convince them that the jihad was succeeding. IS-K tried to diminish the Taliban’s achievements and to stimulate feelings of revenge, for example by claiming that the Taliban had deliberately killed family members of IS-K members during their raids on city cells.

Initially the Taliban were taken aback by the dramatically expanded output of IS-K’s rather slick propaganda. The GDI responded by targeting IS-K activism on social media, exploiting the recruitment efforts of IS-K to infiltrate its own agents, and succeeding in capturing some online activists. It also managed to seize control of some accounts linked to IS-K, and to develop more effective counter-propaganda. A key theme of Taliban propaganda, distributed through the regime’s media as well as on social media, was to portray IS-K as heretics. A pro-Taliban a’lim argued that IS-K members “should be treated like khawarij [heretics] and their Sharia sentences should be hanging or beheading”. Another a’lim argued that IS-K members “are all khawarij” and that the doctrine is clear that under Islamic law, the punishment for this is death. Overall, however, at the end of 2022 online propaganda was the only domain in which IS-K dominated.

The Overall Impact on IS-K in 2021–22

Although IS-K propaganda continually claimed that its numbers were rising, when asked for details, sources provided numbers that in fact showed that the group’s size had remained fairly stable in 2021–22, at just under 8,000 men in total. Most of these in June 2022 were already claimed to be in the north/northeast, according to a source who was himself about to be transferred there from the east.

IS-K sources and propaganda also claimed that recruitment was strong in 2022. When challenged for figures, two IS-K sources provided roughly consistent figures: total new recruitment into IS-K was estimated at 150–200 per month in mid-2022. The main sources of recruits were still identified as “Salafi madrasas, schools, mosques [and] scholars”. As noted elsewhere, IS-K recruitment in universities can be estimated in the low hundreds per year. Overall, these figures seem relatively modest, considering that IS-K was taking losses and suffering defections, and they are consistent with a substantial stagnation in IS-K’s strength during this period.

In sum, IS-K was able to preserve its manpower and appears to have tailored the level and character of its activities to its ability to recruit and, presumably, spend. During this period, however, the Taliban were rapidly expanding their manpower. IS-K’s transition to a fully underground structure had been fairly smooth, with diversions proving rather successful in distracting the Taliban for some months. It is, however, clear that the group had not been able to seize back the initiative and that its financial difficulties seemed to be worsening.

Conclusion

How did the Taliban structure their post-August 2021 mix of tactics for countering IS-K? And how successful were these in fighting the group? Selective violence quickly became the default choice of Taliban policymakers. Identifying the boundaries between extremists, supporting milieus and “quietists” was, however, always contentious. It should also be noted that the Taliban appear to have purposely used bursts of indiscriminate violence to warn hostile populations of what an all-out war with the Emirate would mean for them, and to intimidate them into submission. An aspect of the Taliban’s counter-IS effort that emerges clearly from this paper is that repression, even indiscriminate repression, and reconciliation deals were seen as functional to each other: the stick and the carrot. The new state had to show that it meant business, and that it was able to impose intolerable suffering on the Salafi community if it refused to collaborate.

IS-K’s leadership appears to have underestimated the ability of the Taliban to adapt quickly. Taliban intelligence, despite some obvious limitations, was able to quickly establish a wide and thick network of informers. As insurgents, the Taliban had had a well-developed intelligence network, and they adapted this; they also seem to have prioritised investment in their intelligence agency. Given IS’s reputation for ruthlessness, it was easy for them to obtain the cooperation of bystanders. At the national leadership level, there seems to have been an understanding of the risk of getting trapped in a cycle of violence, and there were interventions to contain the excesses of provincial officials, especially as the new security apparatus consolidated. The Taliban showed their ability to adapt by developing the sophisticated means to make selective repression viable, for example through setting up social media infiltration teams. Still, when selective repression proved difficult to implement because of insufficient intelligence, local Taliban officials usually had no qualms about reverting to indiscriminate violence, even if the scale never approached the main wave of violence of autumn 2021. It is noteworthy in this regard that the Taliban failed to apply the rule of law to counter-IS efforts. The system remained prone to abuse even from the standpoint of Islamic law, and avoiding excesses was always dependent on interventions from the higher leadership levels.

The Taliban also tentatively began working at local reconciliation deals with Salafi communities, but the effort was weakly supported by Kabul and, as of early 2023, it was poorly followed up. National-level talks with the Salafi ulema helped the Taliban shift away from indiscriminate violence, but did not lead to any progress towards an elite bargain. The Taliban were offering peace to the Salafis as subjects of the Emirate, but the Salafi ulema were seeking inclusion.

Where the Taliban were most effective was with choking-off tactics, constraining the ability of IS-K to recruit, resupply and keep money coming in. They waited until they had sufficient manpower available before mounting large-scale military sweeps, to be able to hold the ground afterwards. If they had been engaging in ineffective sweeps, as the previous regime had, they would have alienated the population for no gain.

A pertinent question is how much of the Taliban’s counter-IS effort has derived from their previous experience as insurgents. While none of the sources directly commented on this point, it seems likely that their reluctance to engage in big military sweeps might derive from having experienced the ineffectiveness of such tactics when they were on the receiving end of them before August 2021. Similarly, having had to recruit new insurgents for 20 years, the Taliban seem well aware of the greater difficulties that an insurgent organisation faces when it lacks territorial control. The Taliban furthermore always argued that the indiscriminate revenge-taking and repression practised by Afghan and US security forces in 2001–04 drove many into their ranks, enabling them to start their insurgency. In the current case, however, they have struggled to implement a coherent policy of selective repression, showing perhaps that learning lessons could well be disrupted by the emotional legacy of a long war. Another example of how hatred for the enemy gets in the way of rational policymaking is the Taliban’s failure to follow up on their good start on reconciliation and reintegration.

IS-K undoubtedly proved a resilient organisation after August 2021. Despite facing morale and financial issues, it focused on an urban strategy while trying to strengthen its positions in northern Afghanistan. Militarily speaking, it did not mount a serious threat to the Taliban. The leadership opted to spare its fighters, soon even giving up early attempts to wage a guerrilla war in the east. IS-K tried instead to keep the Taliban busy guarding the cities against a massive wave of urban terrorism, while at the same time expecting its efforts to establish itself firmly in the north to be bearing fruit in the medium term. Time, however, was not on IS-K’s side, and the group’s financial difficulties only increased during 2022.

IS-K appeared to be in a corner by the end of 2022 and early 2023, in good part due to Taliban efforts to counter it. The organisation was surviving by keeping a very low profile, but this meant limited recruitment opportunities and, importantly, far too little fundraising inside Afghanistan. The dependence on money coming from abroad was increasingly proving a liability during 2022. Without financial resources, IS-K was not well positioned to exploit the Taliban’s remaining vulnerability: the fact that the Salafi community, while in general acknowledging a reduction of the pressure exercised by the Emirate, still feels oppressed and very pessimistic about its future under the new regime.

It seems clear that IS-K was very vulnerable to the reconciliation efforts deployed by the Taliban, and that a decisive defeat of the organisation could have been achieved if the Taliban had followed through and implemented their reconciliation packages consistently. Instead, as the IS-K threat appeared to be receding in the second half of 2022 and Taliban self-confidence grew, reconciliation efforts lost steam, despite evidence suggesting that this was the most effective path. It was assumed that defectors would easily reintegrate with the help of the community elders, who, however, received no support from the Emirate. The main reasons for this appear to have been animosity against IS-K within the Taliban’s ranks, fuelled by the considerable amount of blood spilled; resentment over the allocation of scarce financial resources to paying reconciled opponents; and the failure to make significant progress towards a wider elite bargain involving Salafi elites.

Time will tell if the failed reconciliation process is going to be a great missed opportunity for the Taliban. IS-K’s financial weakness could lead to its terminal decline without much Taliban effort, of course, but financial difficulties could still be reversed in the future, in which case the Taliban might regret having neglected their promising reconciliation efforts. While the strong foreign component of IS-K is clearly not susceptible to being enticed to reintegrate, IS-K nowadays needs Afghan participation more than ever – it cannot rely on Pakistanis for dispersed underground operations in cities and villages. If the Taliban were able to substantially cut into IS-K’s approximately 3,000 Afghan members, the group’s viability as an insurgent organisation in Afghanistan would be comprehensively undermined.

Antonio Giustozzi is the senior research fellow at RUSI in the Terrorism and Conflict research group. He has been working in and on Afghanistan in various respects since the 1990s and has published extensively on the conflict and specifically the Taliban and the Islamic State. His main research interests are global jihadism in Afghanistan, Pakistan and Iran, the security sector, state-building and insurgencies. He is currently project director for Strive Afghanistan, which is pioneering new P/CVE approaches. He is also associated with the LSE (South Asia Centre) and was previously associated with War Studies at KCL.