These aren’t AI firms, they’re defense contractors. We can’t let them hide behind their models - The Guardian
Summary
The article argues that major AI firms are functioning as defense contractors, deeply integrated into modern warfare's targeting architecture, rather than neutral technology providers. It highlights the use of AI systems in recent conflicts, such as Israel's war in Gaza and the US-Israeli Iran war, where AI-driven targeting has led to significant civilian casualties, including the strike on the Shajareh Tayyebeh elementary school in Minab. The author contends that AI systems automate a "chosen blindness," creating deniability and making violence appear inevitable, while systematically dismantling accountability frameworks in international humanitarian law. Companies like Palantir, Google, Amazon, Microsoft, Anduril, and OpenAI are identified as key suppliers, with their systems generating targets at speeds impossible for human teams, often with minimal human oversight. The piece criticizes the current regulatory landscape, noting that existing laws and proposed AI acts largely exempt military applications, allowing these firms to operate without sufficient oversight or liability.
Key takeaway
For CTOs and executives in the defense sector, your companies are increasingly viewed as defense contractors, not just tech providers. You should anticipate and prepare for stricter regulatory frameworks, including export controls, congressional oversight, and liability extensions up the supply chain. Ignoring these shifts risks significant legal exposure and reputational damage as international courts begin to address corporate liability for AI-assisted unlawful strikes.
Key insights
AI systems in modern warfare automate "chosen blindness," eroding accountability and enabling rapid, opaque targeting decisions.
Principles
- AI targeting systems inherit and scale existing military logics.
- Accountability frameworks are structurally irrelevant in AI warfare.
- Dual-use technology requires specific military regulation.
Method
AI systems process vast data (phone records, movement, social connections) to infer combatant probabilities, generating ranked target lists for rapid, often 20-second human approval.
In practice
- Regulate AI firms as defense contractors.
- Implement export controls on dual-use AI systems.
- Demand explainable AI reasoning for targeting.
Topics
- AI Warfare
- Algorithmic Targeting
- Military AI Ethics
- Defense Contractors
- AI Regulation
Best for: CTO, Executive, Investor, AI Ethicist, Policy Maker, Legal Professional
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by artifical intelligence via Google News.