Skip to content
General Sir Richard Barrons, former Commander of the UK’s Joint Forces Command, examines how AI is reshaping conflict—from suspected misuse of commercial AI tools to the development of “digital kill webs” connecting sensors, data and military action.
AdminSep 11, 20262 min read

AI Is Already Changing Warfare. Can Human Control Keep Pace?

Artificial intelligence is reshaping how conflicts are fought, how targets are identified and how military decisions are made. For governments and defence leaders, the challenge is already here.

In the interview, General Sir Richard Barrons, former Commander of the UK’s Joint Forces Command, describes a transformation reaching across the battlefield and beyond it. Data, AI, robotics, autonomy and bioscience are changing the character of warfare and raising urgent questions about control, accountability and the speed of decision-making.

As Barrons puts it: “The future is now.”

What the reported misuse tells us

The discussion follows an Anthropic report describing suspected attempts to misuse its AI systems for activities including biological weapons research and conventional weapons development.

Examples discussed in the interview include suspected efforts in Yemen to support missile and rocket development, and a Russian group’s reported work on a drone swarm intended to select and attack targets without human involvement.

These are reports of concerning activity, rather than proof that functioning weapons were produced. Nevertheless, they illustrate the breadth of potential misuse facing AI providers and security organisations.

Barrons argues that the cases identified by one company are likely to represent only part of a wider challenge. Both states and non-state actors can seek to apply AI to military operations, cyber attacks and information manipulation.

Crucially, they do not necessarily need the most advanced technology available. A less capable system may still be sufficient for their purposes.

Understanding the “digital kill web”

One of the most significant developments Barrons describes is the digital kill web: a network connecting sensors, data, decision-making and military capabilities.

Information can come from satellites, ships, mobile phones and open-source material. AI helps process that information, identify patterns and support the prioritisation of potential targets.

The network then connects those assessments with possible actions, spanning conventional weapons, cyber operations and activity in the information environment.

The strategic significance lies in the connection between these elements. Gathering information, assessing it and acting on it become parts of a much faster, more integrated process.

Barrons describes versions of these networks as already operating in contemporary conflicts. Their continued development places increasing pressure on the time available for human judgement.

Speed creates a challenge for human control

Faster processing can help military organisations respond to complex situations. But the ability to generate a recommendation quickly does not establish that the recommendation is accurate, appropriate or justified.

The central question is whether people retain meaningful authority over decisions to use force.

Barrons highlights the importance of a human sanctioning the allocation of targets to weapons. As systems become faster and more autonomous, that responsibility needs to remain substantive: people must be able to assess, challenge and stop an action.

Human oversight depends on more than having someone nominally involved in the process. It requires sufficient understanding, reliable information and time to exercise judgement.

What this means for defence leaders

The discussion points to a challenge that extends beyond acquiring new technology.

Defence organisations need to consider how AI changes command arrangements, training, accountability and the relationship between people and machines. They also need to understand how adversaries may use widely available systems, including outside conventional military operations.

For governments, the task is to keep policy and oversight relevant as capabilities develop. For industry, it is to identify and respond to misuse while recognising the limits of what any single provider can see.

Barrons’ central message is clear: AI-enabled warfare is a present reality. The pressing question is how to preserve responsible human decision-making as its speed, reach and complexity increase.

RELATED ARTICLES