AI-Enabled Drones and the Quest for a Moral Code

Former GCHQ director David Omand has reversed his earlier stance and now argues that autonomous weapons must be equipped with a programmable moral framework. He says the rapid pace of modern combat makes “human-in-the-loop” control increasingly untenable and urges immediate work to embed ethical parameters in future AI-powered drones.

The Guardian

The Ukraine war and recent Iranian-theatre strikes have shown drones operating at speeds that outstrip human decision cycles. Generative-AI breakthroughs, exemplified by large-language models, have prompted defence planners to consider “on-the-loop” supervision, where humans set mission parameters but machines execute targeting decisions in real time.

The Guardian

• David Omand – former GCHQ chief, now a visiting professor at King’s College London, advocating “adaptive moral control layers.”

The Guardian

• Al Carns – UK armed forces minister, supporting the option to remove humans from the decision loop when required.

The Guardian

• Mustafa Suleyman – chief executive of Microsoft’s AI division, contending that AI cannot be moral agents.

The Guardian

• Chris Cole – director of Drone Wars UK, warning that AI lacks judgment.

The Guardian

• Zee Talat – University of Edinburgh scholar, emphasizing the probabilistic nature of AI.

The Guardian

• U.S. defence establishment – allocating $54 billion in the 2027 budget to “autonomous and remotely operated systems,” including a “drone dominance” programme.

The Guardian

• U.S. 2027 defence budget earmarks $54 bn (£40 bn) for autonomous air, land, sea, and under-sea systems.

The Guardian

• More than 100 startups across the United States and Europe are developing drone hardware and software platforms, ranging from low-altitude surveillance craft to armed heavy drones.

The Guardian

• Companies such as Palantir and Anthropic supply AI tools that shorten the “kill chain” in ongoing operations.

The Guardian

Proponents claim that a rigorously coded ethical layer could outperform stressed human operators, reducing collateral damage in split-second engagements. Opponents argue that codifying international humanitarian law into binary code risks oversimplifying complex, culturally contested moral judgments and may create legal vacuums regarding accountability for unlawful strikes.

Former GCHQ Chief Demands Moral Code for AI Weapons

Omand asserts that AI can now evaluate target legitimacy, civilian risk, and identification accuracy, proposing a six-variable framework adjustable by a supervising human. The UK defence policy stipulates “context-appropriate human involvement” in weapons that identify, select, and attack targets, allowing parameter setting by operators. The U.S. budget reflects a strategic shift toward “drone dominance,” yet former CIA director David Petraeus notes the absence of a formal doctrine for autonomous formations.

The Guardian

Drone Wars UK’s Chris Cole contends AI “simply not capable of making a judgment,” while Zee Talat warns that probabilistic models cannot capture the open-ended, debate-based nature of morality. Jessica Dorsey of Utrecht University highlights the difficulty of determining whose moral code a drone should follow, noting the lack of a global consensus on autonomous weapons governance.

The Guardian

Sources diverge on feasibility: Omand and industry advocates claim current AI can replicate human targeting reasoning, whereas academics and anti-drone activists argue that moral reasoning remains uniquely human and that legal frameworks lag behind technological capability. No source provides empirical evidence of fully autonomous lethal strikes adhering to international law.

The Guardian

• “My call is to really get some work done on this, so that we’re not left in a situation where there isn’t a moral component built into future AI-powered weapon systems,” — David Omand, former GCHQ chief

The Guardian

• “Speaking to the Guardian before an appearance at the Cheltenham science festival on Wednesday, Omand said AI technology was now capable of weighing the factors that go into a human drone operator’s targeting decisions, such as whether a target was legitimate, whether there would be civilian casualties and whether the target has been correctly identified.” — David Omand

The Guardian

• “you must have the ability to take the human out of the loop when required” — Al Carns, UK armed forces minister

The Guardian

• “AIs cannot be people – or moral beings.” — Mustafa Suleyman, Microsoft AI chief

The Guardian

• “AI is simply not capable of making a judgment. It merely processes data, completely lacking the ability, for example, to distinguish civilians from combatants or to judge whether loss of life is proportionate to military advantage.” — Chris Cole, Drone Wars UK

The Guardian

• “If you have a machine that’s probabilistic by nature it will veer towards the most likely answer in a situation. Do we think that morality follows probabilistic notions?” — Zee Talat, University of Edinburgh

The Guardian

UK minister Al Carns and U.S. defence planners are urging development of “on-the-loop” capabilities, while scholars and NGOs call for a binding international treaty to regulate autonomous weapon morality. Ongoing commissions, such as the 2014 University of Birmingham review chaired by Omand, and emerging multilateral discussions at the United Nations signal that the debate will shape future arms-control negotiations.

The Guardian