Top AI developers including OpenAI, Anthropic, Google, and Microsoft have issued a joint open letter warning that advanced AI could soon enable sophisticated cyberattacks. They stress a "limited window" to bolster defenses, noting AI-powered attacks rose 89% in 2025. The letter urges investment in security teams, AI-driven defense tools, and information sharing.

It calls for upgrading vulnerable systems, continuous testing, and government support to secure digital infrastructure.