Use of AI in war: Ethics under fire

THE unprovoked US-Israel war against Iran entered its fourth week. The most tragic image associated with Operation Epic Fury, is the 'accidental' death of over 175 Iranians, mostly schoolgirls, due to a US missile attack in the early stages of this preemptive war launched on February 28.Open source intelligence confirms that the ordnance used was a Tomahawk land-attack cruise missile fired from a US naval platform and it struck a primary school for girls at Minab in southern Iran.Why was a primary school struck in the first phase of the war? More so, when Israel has demonstrated the ability to eliminate pinpointed individuals and choose vital targets, with considerable attention paid to their military relevance in the first attack. Was AI (artificial intelligence) a contributing factor for this unforgivable slaughter?Preliminary investigation by the US Pentagon suggests that this was a "targeting mistake" due to outdated intelligence that showed the site as a vital military complex, being fed into the operational plan. The school had previously been part of an IRGC (Islamic Revolutionary Guard Corps) complex, but had functioned as a civilian site for nearly a decade. Evidently, the school was mistakenly identified as a military site and retained on the US target list. This is inept intel collection and reinforces the image of the US military as being trigger-happy, when it comes to 'mistakenly' killing civilians — as was the case in Afghanistan.The Pentagon has initiated a high-level investigation and the preliminary assessment is that while AI was not directly responsible for the deaths of the schoolchildren, its use in the target selection process was a contributing factor.AI has been used extensively in the current war and has contributed both to the narrative in which the hostilities have been framed by either side and the actual military operations.The narrative part includes industrial-scale disinformation and fake imagery comprising AI-generated images, videos and fabricated satellite imagery. This content has inundated social media platforms, making it difficult to differentiate between the real and the manufactured content.Further, AI has been used to create deepfakes of war destruction by generating videos of missile strikes, such as make-believe explosions. Fabricated satellite imagery and a "troubling tactic", whereby AI-enabled fake, technical-looking analysis was disseminated to falsely discredit authentic, real-world evidence of the war was part of the fake news deluge that added to the fog of war and, very often, real facts remained blurred.AI-generated propaganda, often malicious and vicious, was created in ample measure and this included both anti-US and anti-Semitic posts and the demonising of Iran and the 'mad mullahs'. AI chatbots were innovatively exploited, often identifying AI-generated fakes as real news.This surge in fake or embroidered AI content has resulted in "information pollution", making it difficult to distinguish facts related to the actual war against Iran and the fiction embedded in the 'war over reality'.Concerns about the military misuse of AI stem from the perceived 'rubber stamping' of targets - namely, that the US and Israel, with their proven domain competence in this niche, are using AI to identify, prioritise and target specific assets/positions, thereby leading to a dangerous reduction in human oversight. This is termed rubber-stamping of automated recommendations.The rush to use AI for targeting and prioritising speed over deliberation can result in high numbers of civilian casualties and the Tomahawk strike on the Minab school is illustrative.As per reports from Washington, the US used Anthropic's AI tool Claude to identify and strike around 900-plus targets in the first 24 hours of Epic Fury and, evidently, the Minab school was one of the unlucky sites in the AI-generated list.At which stage was there a human intervention before the order to 'kill'/ strike was given? This will be crucial to assigning human responsibility for authorising the killing of the adversary with the help of AI and a host of legal and ethical issues loom large.The core concern about AI in warfare is the potential lack of human oversight in life-and-death decisions and the three major conceptual challenges are complex:a) Autonomy vs accountability: as AI systems gain decision-making power, it becomes challenging to hold anyone in the human ladder accountable for mistakes or unethical actions.b) Absence of human judgment: AI might not fully grasp complex contexts, leading to unintended consequences or disproportionate responses.c) Escalation risks: autonomous systems could accelerate conflict escalation, outpacing human intervention.In summary, the integration of AI directly into the targeting process has dramatically reduced the time factor in the 'kill-chain'. Notably, AI was used in a rudimentary manner by the US in the 1991 war for Kuwait — that was more for logistics. In the intervening decades, technology has made AI far more capable and potentially lethal on the battlefield.Currently, there is no specific international treaty or binding law that bans or comprehensively regulates lethal autonomous weapons systems (LAWS), often called "killer robots" — weapons that can select and engage targets without meaningful human control. This is akin to video games and the greatest danger is that actual killing will become remote and contactless 'fun' — a turn of phrase that President Donald Trump used in relation to the sinking of the Iranian warship, the IRIS Dena.The UN Convention on Certain Conventional Weapons (CCW) seeks to restrict the use of weapons causing unnecessary suffering or indiscriminate effects and a Group of Governmental Experts (GGE) on LAWS has met regularly to address challenges.The Seventh Review Conference of the CCW, scheduled to meet in November, is expected to decide the next steps, potentially launching formal negotiations for an international protocol or a treaty that will regulate the use of AI in war.But when the most powerful nation in the world has declared that it is not bound by any law and embarked on Epic Fury in a reckless manner without the customary "rules of engagement", it would be an epic folly to expect that ethics and humanism will temper the advent of AI.

Use of AI in war: Ethics under fire

The core concern about AI in warfare is the lack of human oversight in life-and-death decisions.

C Uday Bhaskar

Use of AI in war: Ethics under fire

The core concern about AI in warfare is the lack of human oversight in life-and-death decisions.

C Uday Bhaskar

Unlock Premium Insights in This Article