Google removes its pledge not to build weapons
On 4 February 2025 Google updated its AI Principles, removing the list of applications it would not pursue, weapons and surveillance among them. In their place came language about developing AI where "the likely overall benefits substantially outweigh the foreseeable risks".
Why it matters
For seven years the most cited corporate boundary between AI and the use of force rested on a list of four lines. It was removed in a day, which showed what a voluntary limit is worth when it is written nowhere but on the company's own page.
The page had said Google would not design or deploy AI in four application areas, and all four went: technologies that cause or are likely to cause overall harm; "weapons or other technologies whose principal purpose or implementation is to cause or directly facilitate injury to people"; "technologies that gather or use information for surveillance violating internationally accepted norms"; and "technologies whose purpose contravenes widely accepted principles of international law and human rights". The new text promises "appropriate human oversight" and alignment with "widely accepted principles of international law and human rights". International law survives as a bearing, then, rather than as a list of what the company will not do. The change came with a blog post by James Manyika, senior vice-president, and Demis Hassabis on Google's 2024 responsible AI report; the post does not mention the removed list. It is worth noting who did not sign it: the original 2018 principles were published by the chief executive, Sundar Pichai, in person.