US and Russia strip AI weapons safeguards from UN pact

Negotiations regarding Lethal Autonomous Weapons Systems (LAWS) under the UN Convention on Certain Conventional Weapons (CCW) have historically been characterized by a divergence between states seeking binding prohibitions and major military powers advocating for non-binding guidelines. Since 2014, discussions have frequently stalled over definitions of 'meaningful human control' and the scope of applicability to dual-use technologies. Major AI-capable states, including the US, Russia, and China, have consistently resisted treaty language that would constrain operational flexibility or mandate specific technical architectures for target verification. This pattern reflects a broader precedent in arms control where emerging technology regimes often face delays due to verification challenges and strategic ambiguity. Previous CCW review conferences have resulted in political declarations rather than legal instruments when consensus among major powers was absent. The recurring dynamic involves civil society organizations pushing for preemptive bans while state actors prioritize retaining optionality in rapidly evolving threat environments.

On September 28, 2026, Human Rights Watch (HRW) issued a public statement alleging that delegations from the United States and Russia collaborated to remove critical ethics and human-oversight safeguards from a draft United Nations pact on lethal autonomous weapons systems (LAWS). According to HRW, this removal occurred during a fifteen-hour marathon editing session initiated by US and Russian representatives on September 27, 2026, just prior to a scheduled UN review conference. The organization claims that these last-minute revisions systematically stripped provisions requiring human review of AI-selected targets, as well as clauses related to predictability and ethical constraints, effectively gutting three years of multilateral negotiation.

Multiple news outlets, including IBTimes, The Mirror, MSN, Times Now, Crypto Briefing, and New Kerala, reported on the allegations, citing HRW as the primary source. These reports describe a scenario in which American and Russian diplomats worked in concert during the final hours of talks in Geneva to weaken regulations governing what critics term 'killer robots.' According to Crypto Briefing, the two nations are the most invested in military AI and allegedly teamed up to ensure the framework would not impose binding restrictions on their respective programs. Times Now characterized the process as 'death by a thousand paper cuts,' suggesting a deliberate strategy to erode the treaty's efficacy through incremental textual changes rather than outright rejection.

The United States delegation has not publicly addressed the specific allegations made by Human Rights Watch regarding the September 27 editing session. However, the US has historically opposed binding international restrictions that limit military flexibility in AI development, favoring non-binding norms or political declarations over enforceable treaties. Similarly, the Russian delegation has not commented on the alleged edits but maintains a longstanding position resisting international constraints on autonomous weapons development. Neither party has confirmed or denied the specific claim that they coordinated to strip safeguards during the fifteen-hour window.

The controversy centers on the tension between technological sovereignty and international humanitarian law. Critics argue that removing human oversight provisions risks normalizing fully autonomous lethal targeting, where algorithms select and engage targets without meaningful human control. Defenders of the current approach, while silent on this specific incident, have previously argued that premature binding regulations could stifle innovation and fail to account for the strategic realities of modern warfare. The timeline indicates that formal UN autonomous weapons negotiations began on September 1, 2023, making the alleged eleventh-hour changes particularly contentious given the duration of prior diplomatic efforts.

Human Rights Watch asserts that the removed safeguards were essential for maintaining accountability and adherence to international law. Their statement implies that the revised draft no longer guarantees that humans will remain in the loop for lethal decisions. Conversely, the absence of a public rebuttal from the US or Russian delegations leaves the factual basis of the coordination claim unverified by the accused parties. The narrative currently rests entirely on NGO reporting and secondary media amplification, with no official treaty text or diplomatic record yet released to confirm the extent of the alleged deletions.

The controversy implicates global humanitarian law standards and the future of multilateral arms control. If safeguards are indeed removed, states developing lethal autonomous systems may operate without internationally agreed human-review thresholds, potentially increasing civilian risk and lowering the barrier to automated conflict escalation. Affected parties include civilian populations in future conflict zones, diplomatic communities relying on CCW frameworks, and defense industries navigating regulatory uncertainty. While no monetary fines or job losses are at stake, the normative infrastructure governing military AI faces potential degradation. The magnitude is systemic rather than quantifiable in economic terms, hinging on whether the November UN review conference adopts the weakened text or demands restoration of oversight provisions.