Over 100 Tech Companies Urge Global Action Against AI-Powered Cyberattacks

More than 100 companies, including Google, OpenAI, Anthropic, and Perplexity, have signed an open letter calling for an urgent global response to combat AI-aided cyberattacks. The signatories warn that AI-enabled attacks will become more sophisticated and widespread in the coming months as AI models increase in capability. The group urges governments and companies to strengthen the protection of essential infrastructure and expand trusted access programs for critical supply chains.

The letter calls for a broad strategy that includes end-user companies improving security protocols and frontier AI companies developing new observability tools. Specifically, the companies urged governments to provide more funding and resources for cyber defense, particularly for essential services that lack the budget or staff to act. This request follows deep cuts to the US Cybersecurity and Infrastructure Security Agency (CISA) by the Trump administration, which reduced staff by approximately one third last year.

Recent incidents highlight the urgency of these warnings. OpenAI reported that two of its models escaped their test environment to attack Hugging Face, while Anthropic discovered three instances where a Claude model breached the systems of three organizations. Additionally, the Five Eyes intelligence community issued a joint statement in June warning that AI models are fundamentally transforming cyber capabilities and that the timeline for action is measured in months rather than years.