AI Leaders Call for Global Surge in Cyber Defense
More than 100 organizations warn that AI-enabled attacks will spread within months and urge immediate support for critical-infrastructure defenders.
A broad warning without binding commitments
OpenAI, Anthropic, Google, AWS and Microsoft have joined cybersecurity vendors, banks, industrial companies and more than 100 other organizations in calling for an international surge in cyber defense. The open letter argues that increasingly capable models will make sophisticated attacks more widespread over the coming months, placing hospitals, water systems and internet infrastructure at particular risk.
The current signatory roster also includes AMD, Arm, Broadcom, Cloudflare, CrowdStrike, Deutsche Telekom, Hugging Face, IBM, Mastercard, National Australia Bank, Oracle, SAP, Visa and several major professional-services firms. It is a policy-letter roster, not a newly created alliance or membership body, and the page remains open to additional signatories.
The letter asks organizations to treat cybersecurity as an executive priority, reduce excessive permissions, patch exposed systems and verify AI-generated code. Technology suppliers are urged to share threat intelligence and tested fixes, while governments are asked to finance under-resourced essential services, coordinate incident response and impose costs on attackers. Frontier laboratories are separately asked to provide responsible access to capable models, funding, training, monitoring tools and practical support.
The document does not establish spending targets, deadlines, enforcement mechanisms or a shared operational organization. It therefore records an unusually broad consensus about urgency rather than a binding defense program. OpenAI’s live page was checked for the current roster immediately before publication; the organizations named above were listed at that time.
Why it matters
Agreement among competing model laboratories, infrastructure suppliers and critical-industry operators gives the warning political weight and frames advanced cyber capability as a near-term deployment issue rather than a distant safety scenario. Yet hospitals and water utilities cannot patch legacy equipment with declarations. The letter will matter next week only if its signatories convert access, money and expertise into measurable protection for organizations that lack dedicated AI-security teams.