The Pentagon banned Anthropic's Claude AI from military operations, marking the first foundation model restriction by U.S. defense authorities. The decision follows leaked internal Anthropic memos questioning military applications and Constitutional AI safeguards for weapons development.
India launched a $15 billion sovereign AI program targeting independence from U.S. and Chinese foundation models. The initiative funds domestic model development and computing infrastructure as nations pursue strategic autonomy in AI capabilities.
Seven countries now deploy foundation models for military reconnaissance, defense analysts confirmed. Applications include satellite imagery analysis and logistics optimization. No systems make autonomous weapons decisions under current international law.
Mistral CEO Arthur Mensch framed AI competition as open versus closed systems rather than geographic rivalry. "Power concentration in AI development poses greater risks than where systems are built," Mensch said, advocating open-source models to distribute control.
Geopolitical tensions escalated after the U.S. restricted GPU exports to 23 countries citing weapons proliferation risks. China retaliated with rare earth mineral export controls critical for chip manufacturing. The European Union proposed AI sovereignty requirements for government contracts.
Lawsuits against AI companies over safety failures surged 340% in 2025. Cases cite inadequate testing and harmful outputs. Courts examine whether foundation model developers bear liability for downstream applications.
Three major AI labs employ former defense officials in leadership roles, raising transparency concerns. Advocates demand disclosure of military partnerships and dual-use research amid revolving door dynamics.
The crisis exposes control questions over transformative AI systems. International AI safety summits scheduled for April target coordination mechanisms, though binding agreements remain unlikely as technical progress outpaces governance frameworks.