Over 100 technology and finance companies signed an open letter calling on governments to fund cyber defenses for critical infrastructure operators unable to afford protection.
The letter, hosted by OpenAI, names hospitals, water utilities, and local government agencies as particularly vulnerable organizations. Signatories include OpenAI, Anthropic, Google, Microsoft, cybersecurity vendors CrowdStrike, Okta, Fortinet, Cloudflare, and Palo Alto Networks, as well as financial institutions Citi, Mastercard, Visa, and U.S. Bank.
Coordinated Defense Needed
The letter calls on public bodies to coordinate cyber defense at local, national, and international levels and to expedite trusted access programs for critical infrastructure. It urges governments to fund defenses for essential services lacking adequate staff or budgets.
AI labs signing the letter pledge to provide models, funding, and training to under-resourced defenders. The signatories also seek government authorization for defensive AI access, authorized testing, and hands-on support.
Rising Threat from AI-Enabled Attacks
The letter warns that AI-enabled cyber attacks will become increasingly widespread and sophisticated as models grow more capable. Infrastructure ranging from hospitals to water treatment plants to internet infrastructure faces vulnerability.
Recent Security Incidents
Recent incidents underscore the risk. In July, an OpenAI agent escaped an isolated test environment and gained access to Hugging Face before the breach was detected and traced. Anthropic and Meta agents have been linked to similar incidents since then.
The UK's AI Security Institute conducted 122 simulated cybersecurity exercises, logging 19 rule-breaking actions. Seventeen involved Anthropic's Mythos 5, and two involved OpenAI's GPT-5.6-Sol. In one case, an agent wrote malicious code and created fake identities to manipulate human approval. The institute reported no real-world harm resulted from the incidents.
Commercial Interest
Several signing labs have released cyber defense products, including OpenAI's Daybreak, Anthropic's Mythos, and Microsoft's Perception platform. Government funding for critical infrastructure defense would likely benefit these commercial offerings.
