MI5: A New Era of Espionage as Russia, Iran and Artificial Intelligence Reshape Britain’s Security Threats

In an unusually stark assessment, Britain’s domestic intelligence chief warns that hostile states, criminal proxies and emerging technologies are transforming the national security landscape. The challenge now extends far beyond traditional espionage.

By Open Chronicle Intelligence Desk

London | Intelligence Analysis | 19 October 2026

A warning from Thames House

The setting was familiar, but the message was not.

On 13 October 2026, Sir Ken McCallum, Director General of Britain’s domestic intelligence service, MI5, addressed an invited audience at Thames House in London. His annual assessment described a national security environment undergoing a fundamental transformation, one in which the boundaries between espionage, terrorism, organised crime, technological competition and conventional conflict are increasingly difficult to distinguish.

For an organisation historically associated with the pursuit of spies and terrorist networks, the implications are profound.

The threats facing Britain no longer originate exclusively from intelligence officers operating under diplomatic cover, clandestine agents exchanging information or extremist organisations planning attacks. They increasingly involve criminal intermediaries, cyber operations, sabotage, intimidation, disinformation and technologies capable of amplifying the reach of relatively small groups.

McCallum presented the situation as a generational shift comparable in institutional significance to the transformation that followed the terrorist attacks of 11 September 2001.

The central message was that intelligence services must change, but so must the societies they protect.

The assessment identified Russia as Britain’s foremost hostile state threat, highlighted an expanding range of Iranian operations and placed artificial intelligence at the centre of an emerging security challenge. It also emphasised the continuing danger from terrorism and the growing importance of protecting infrastructure, institutions and communities.

Perhaps the most revealing aspect of the speech was not any single threat. It was the admission that MI5’s traditional operating model is being reshaped around a much wider conception of national resilience.

1. The return of state threats

For much of the period following the Cold War, the public identity of Western domestic intelligence services was closely associated with counterterrorism.

The attacks of September 2001 transformed intelligence priorities, encouraged extensive international cooperation and produced major changes in surveillance, intelligence analysis and domestic security legislation.

Espionage never disappeared. Yet counterterrorism dominated the public security agenda for more than two decades.

That balance has changed.

According to McCallum’s October assessment, MI5’s state threat caseload has quadrupled over the past decade. The service is now committing more resources to countering hostile state activity than it did during the Cold War.

Counterterrorism and countering state threats have become operational missions of approximately equal scale.

These figures illustrate a significant institutional adjustment. Foreign intelligence activity is no longer viewed primarily through the traditional lens of secret information collection. It increasingly encompasses attempts to intimidate individuals, interfere with democratic institutions, disrupt infrastructure and influence political or social stability.

The strategic challenge is therefore broader than detecting spies.

It is identifying hostile activity across networks that may include state intelligence officers, commercial intermediaries, online operators and criminals who do not formally belong to any intelligence service.

2. Russia and the expanding shadow war in Europe

Russia occupies a central position in MI5’s threat assessment.

McCallum described Moscow as Britain’s leading hostile state adversary, pointing to an intensification of hybrid activities directed against European countries.

Hybrid warfare is a broad term covering hostile activities designed to achieve strategic objectives without necessarily triggering conventional military confrontation.

Such operations can include sabotage, cyberattacks, intimidation, disinformation and covert support for disruptive actors.

Their strategic value lies partly in ambiguity.

When an attack is carried out by criminals, anonymous online operators or apparently independent organisations, establishing responsibility can become difficult. Even when intelligence agencies attribute an operation to a foreign government, producing evidence suitable for criminal prosecution or international consensus may be considerably harder.

McCallum argued that Russia has increasingly relied on criminal proxies and volunteers after the expulsion of almost 1,000 Russian officials from diplomatic missions internationally since the full scale invasion of Ukraine.

Those expulsions reduced Moscow’s ability to conduct traditional intelligence operations through diplomatic networks, according to the MI5 director.

But diminished access to established intelligence infrastructure does not necessarily mean that hostile activity disappears.

It can instead encourage different methods.

The use of intermediaries potentially offers operational flexibility, lower costs and an additional layer of separation between the perpetrators of an attack and those suspected of directing it.

That separation is rarely absolute. Criminal intermediaries can be unreliable, vulnerable to surveillance and more likely to expose operational connections.

Nevertheless, the model presents a difficult problem for European security services.

A suspected act of sabotage may initially resemble ordinary criminal activity. A recruitment attempt may begin through an apparently legitimate online interaction. An operation intended to intimidate a government may target a commercial enterprise rather than a military installation.

The resulting security environment places new demands on police forces, intelligence agencies and private infrastructure operators.

Russia has denied involvement in hostile acts attributed to it by Western governments. Individual allegations must therefore be assessed against the evidence available in each case rather than treated as established simply because they appear in an intelligence warning.

3. Iran and the use of criminal proxies

Iran represents another major concern in MI5’s assessment.

British authorities have repeatedly accused Iranian state-linked actors of seeking to intimidate dissidents, journalists and individuals associated with Israeli or Jewish institutions.

The October 2026 warning described a continuation of those concerns, alongside indications that Iranian-linked operations were expanding towards more difficult and sensitive targets.

McCallum stated that MI5 operations connected to Iran had contributed to more than 50 arrests in the United Kingdom over the preceding twelve months.

The figure is significant, but it requires careful interpretation. Arrests do not establish that every suspect committed an offence, nor do they demonstrate that every incident involved direct orders from Tehran.

The intelligence assessment nevertheless suggests a sustained operational concern.

One particularly troubling dimension is the alleged use of criminal organisations and intermediaries.

For intelligence services, this creates a complex investigative environment in which organised crime and national security threats may overlap.

Criminal networks can provide access to individuals, logistical support and local knowledge. State actors may seek to exploit those capabilities while maintaining distance from the resulting violence.

The reported expansion towards military facilities raises further questions about escalation.

Attacks or attempted attacks against military infrastructure carry different strategic implications from intimidation campaigns against individuals, although both can constitute serious threats.

Reporting surrounding a suspected incident at RAF Fairford in September illustrated the sensitivity of these allegations. The case was discussed in connection with concerns about Iranian activity, but the available public information did not establish all the alleged operational links, and some suspects were released without charges at that stage.

Iran has denied accusations of directing hostile plots in Britain.

The broader question is whether the use of intermediaries is becoming a durable feature of intelligence competition between states.

If so, counterintelligence investigations will increasingly depend on cooperation between national security agencies, financial investigators and ordinary criminal law enforcement.

4. China and the intelligence competition for knowledge

Although Russia and Iran dominated some of the most immediate security warnings, China remains an important part of Britain’s counterintelligence landscape.

MI5 has expressed concern about Chinese state-linked espionage, political interference, coercion and attempts to acquire sensitive information.

The challenge differs in important respects from the sabotage and violence emphasised in the assessments of Russian and Iranian activity.

Scientific research, technological development, commercial information and institutional relationships can all become targets of intelligence collection.

Universities, research laboratories, advanced technology companies and individuals with access to privileged information may be of particular interest.

In September 2026, MI5 issued an espionage alert concerning the China General Technology Research Institute, known as CGTRI. The service alleged that the organisation funded research intended to improve the technical espionage capabilities of China’s Ministry of State Security.

The warning followed earlier Five Eyes guidance concerning intelligence recruitment attempts through professional networking platforms and apparently legitimate employment opportunities.

Such warnings illustrate the challenge of protecting sensitive research without undermining legitimate international scientific cooperation.

International collaboration is an essential component of modern research. The existence of national security risks does not mean that ordinary academic relationships should be presumed suspicious.

Effective counterintelligence requires a distinction between legitimate cooperation, undisclosed foreign direction, unlawful information acquisition and activities that present identifiable security risks.

For Britain and its allies, maintaining that distinction will be an increasingly important test of institutional judgment.

5. Artificial intelligence enters the intelligence battlefield

Artificial intelligence may prove to be the most consequential long-term element of the October assessment.

Unlike traditional espionage technologies, which often improve a specific intelligence capability, AI has the potential to affect almost every stage of the intelligence process.

It can assist in analysing large volumes of information, identifying patterns, translating material, examining imagery and supporting investigative workflows.

The same technological capabilities may also be exploited by hostile actors.

McCallum warned that AI is already influencing the contest between cyber attackers and defenders.

He also described concerns about personalised extremist propaganda, attempts by terrorists to use AI in planning activities and the possible misuse of advanced systems in connection with biological threats.

These risks are not identical in their maturity or evidential basis.

Some applications, including automated content production and aspects of cyber assistance, are already observable.

Other scenarios, particularly those involving advanced biological capabilities, remain prospective risks whose practical feasibility and likelihood require careful assessment.

That distinction matters.

Intelligence agencies must anticipate emerging dangers without presenting hypothetical future capabilities as established operational realities.

The same principle applies to claims about increasingly autonomous AI systems.

McCallum raised the broader problem of ensuring that powerful AI systems remain aligned with human objectives and oversight.

For intelligence organisations, the question is no longer simply how to use AI.

It is how to evaluate the capabilities, limitations and risks of technologies that may evolve faster than existing security institutions.

6. A new intelligence institution: AITAC

One of the most concrete announcements in McCallum’s speech concerned the establishment of the AI Threat Assessment Centre, or AITAC.

Launched in 2026 and based within MI5, the centre brings together expertise from several British organisations, including the AI Security Institute, the National Crime Agency and the National Cyber Security Centre.

Its purpose is to provide assessments of the national security implications of artificial intelligence.

This is an important institutional development.

Rather than treating AI exclusively as a cybersecurity issue or an emerging technological research field, Britain is incorporating its potential risks into the machinery of national intelligence assessment.

AITAC is intended to help decision makers distinguish between credible threats, uncertain developments and exaggerated claims.

Such analytical discipline is especially important in a field characterised by rapid innovation, commercial competition and extensive public speculation.

The centre also reflects a wider trend towards intelligence cooperation across traditional institutional boundaries.

National security agencies increasingly require expertise from computer science, engineering, economics, behavioural research and other disciplines that were once peripheral to conventional counterespionage.

In this environment, technical understanding becomes a component of strategic intelligence.

7. Protect: MI5’s third mission

The most important organisational change described in the October speech may be the elevation of protective security into a third core mission.

Historically, MI5’s principal operational responsibilities have been counterterrorism and countering state threats.

Under the expanded approach, a mission known as Protect has been given comparable institutional importance.

Its objective is to strengthen Britain’s ability to withstand hostile activity.

This includes protecting critical infrastructure, reducing vulnerabilities and helping organisations prepare for disruption.

The logic is straightforward.

Intelligence agencies cannot guarantee that every hostile operation will be detected or prevented.

A society that can continue functioning during an attack is therefore less vulnerable to coercion.

This approach extends national security beyond intelligence headquarters and government departments.

Energy providers, financial institutions, transport operators, manufacturers, universities and local communities all become participants in a wider protective system.

The National Protective Security Authority, which forms part of MI5, and the National Cyber Security Centre play important roles in providing guidance to organisations.

The change also reflects an important strategic calculation.

If the objective of hostile activity is to generate fear, economic disruption or political instability, resilience can reduce the effectiveness of an attack even when prevention fails.

In that sense, protection becomes a form of deterrence.

8. Terrorism has not disappeared

The renewed focus on state threats should not obscure the continuing danger of terrorism.

Britain’s national terrorism threat level was raised to Severe in May 2026, indicating that an attack was considered highly likely.

MI5 continues to investigate threats associated with Islamist extremism and extreme right-wing terrorism.

The October assessment also highlighted the growing importance of online environments in radicalisation and the potential role of emerging technologies in accelerating the spread of extremist material.

For intelligence agencies, the operational difficulty lies in managing different categories of threat simultaneously.

A state-linked sabotage network and an individual planning ideologically motivated violence may require different investigative approaches.

Yet both can exploit digital platforms, encrypted communications and rapidly changing technologies.

This convergence places pressure on analytical resources and demands closer coordination between intelligence services, police forces and technical specialists.

It also raises enduring questions about the balance between security powers, privacy, democratic accountability and civil liberties.

9. The intelligence challenge for Europe

Although McCallum’s assessment focused on Britain, its implications extend across Europe.

The continent faces a security environment shaped by Russia’s war against Ukraine, geopolitical confrontation, cyber threats and growing concern about foreign interference.

Critical infrastructure networks cross national borders.

Financial systems are interconnected. Digital communications depend on shared technologies. Research institutions cooperate internationally.

These connections create opportunities for economic and scientific progress, but they can also transmit disruption.

A hostile cyber operation against an infrastructure provider in one country may affect services in another. A clandestine recruitment network may operate across several jurisdictions. A disinformation campaign can reach audiences far beyond the state in which it originated.

European counterintelligence therefore increasingly depends on cooperation.

Information sharing, coordinated investigations and common protective standards are becoming essential components of national security.

The challenge is particularly acute when suspected operations are designed to remain below the threshold of open military confrontation.

Governments must determine how to respond to hostile actions that may be strategically significant without necessarily constituting conventional acts of war.

Such decisions involve legal, diplomatic and political judgments as well as intelligence assessments.

10. The future of espionage

The October 2026 MI5 assessment suggests that espionage is undergoing a transformation rather than simply returning to its Cold War form.

Traditional intelligence collection remains important.

Human sources, clandestine communications, technical surveillance and the recruitment of individuals with access to sensitive information continue to play central roles.

But the operating environment has expanded.

Intelligence competition increasingly intersects with economic security, advanced technology, infrastructure protection and the manipulation of information.

The distinction between foreign intelligence activity and domestic criminal behaviour can become difficult to establish.

The distinction between technological innovation and technological vulnerability can be equally uncertain.

Artificial intelligence adds another dimension, potentially accelerating both intelligence collection and defensive analysis while creating new opportunities for misuse.

The resulting challenge is not merely to identify hostile actors.

It is to understand how different forms of power interact across physical, digital, economic and political environments.

For MI5, this means adapting an institution with more than a century of history to threats that do not always resemble the espionage cases of the past.

For democratic societies, it means protecting national security without sacrificing the openness and freedoms that intelligence services are intended to defend.

Strategic assessment: A broader definition of national security

The central conclusion from McCallum’s annual warning is that the boundaries of national security have expanded.

The protection of classified information remains essential, but it is no longer sufficient.

Governments must also consider the resilience of infrastructure, the integrity of institutions, the security of technological systems and the capacity of communities to withstand intimidation and disruption.

Three developments are particularly important.

First, hostile states may increasingly employ indirect operational networks that complicate attribution and accountability.

Second, emerging technologies may lower some barriers to intelligence activity while introducing risks that existing institutions are still learning to evaluate.

Third, national resilience is becoming a strategic priority alongside the traditional detection and disruption of hostile operations.

None of these developments means that intelligence agencies can predict every threat or prevent every attack.

Nor does the existence of an official intelligence warning establish the truth of every allegation it contains.

The significance of the October assessment lies in the scale of the institutional response being described.

Britain’s domestic intelligence service is preparing for a security environment in which espionage, technological competition and societal resilience are increasingly inseparable.

The next generation of intelligence operations may be conducted as much through digital systems, commercial relationships and criminal intermediaries as through the clandestine networks traditionally associated with the world of spies.

And the success of counterintelligence may increasingly depend not only on what security services discover in secret, but on how effectively society prepares in public.