Google AI Researchers Sign Letter Opposing Military AI Contracts
Hundreds of Google AI researchers, including senior personnel from DeepMind, have signed an open letter urging CEO Sundar Pichai to reject a confidential AI contract with the U.S. Department of Defense.
The letter emphasizes that AI systems can concentrate power and are prone to errors, and that Google cannot monitor or limit actual uses under air-gapped confidential networks. It has garnered over 580 signatures (including more than 20 directors and VPs) and was sent to Pichai on Monday.
Market-wise, Google employees and ethical advocates are pushing for a shift in company policy against military AI, leading to a reallocation of funds from potential defense contracts to civilian AI projects and competitors. Google's short-term reputation and government contracts are under pressure, benefiting civilian-focused AI companies like OpenAI and Anthropic.
Source: Public Information
ABAB AI Insight
Google employees previously protested in 2018 against military AI contracts through "Project Maven," forcing the company to withdraw from some agreements. This latest petition with over 580 signatures (including core DeepMind researchers) continues the long-standing internal ethical movement against the application of AI in lethal weapons or mass surveillance.
From a capital perspective, researchers are pressuring Pichai to reject the confidential defense contract, effectively attempting to influence Google's resource allocation to avoid a shift of AI technology towards military use. Their motivation is to maintain the company's "beneficial to humanity" brand image and protect civilian AI talent retention, while also creating a market window for competitors to seize government contracts.
Similar to the aftermath of the 2018 Maven project protests, where Google released AI ethical principles, and the internal disagreements among several tech companies regarding military AI expected in 2025-2026, the current event indicates that Google is at a critical stage of transitioning from civilian AI dominance to facing pressures from dual-use military contracts.
Essentially, this reflects a regulatory shift: the employee petition is transferring the decision-making power regarding military AI from management to internal ethical oversight and public pressure, exposing the responsibility risks of monitoring failures under air-gapped networks, and forcing capital to shift from potentially high-profit defense contracts to civilian and commercial priorities, achieving a structural reconfiguration from government dependency to ethics-driven constraints.