Skip to content
Cardan-AI
Back to news
AI Cybersecurity1 September 2026

116 AI companies — including direct rivals — sign joint call for collective cyber defense

Anthropic, OpenAI, Google, Microsoft, AWS, Cisco, Cloudflare, CrowdStrike, Oracle, Hugging Face and Perplexity are among the 116 companies and organizations that co-signed a call for collective action on August 27, 2026, as AI-enabled cyberattacks grow more sophisticated.

116 technology companies and organizations — including head-to-head competitors in generative AI — co-signed a joint text on August 27, 2026, titled “a call for collective action on cyber defense.” Named signatories include Anthropic, AWS, Cisco, Cloudflare, CrowdStrike, Google, Hugging Face, Microsoft, OpenAI, Oracle and Perplexity, according to CNBC.

The letter rests on three principles: acknowledge that current security practices are insufficient against increasingly sophisticated attacks; equip more defenders — not just large enterprises — with AI-capable cybersecurity tools; and coordinate a unified defensive response rather than scattered individual efforts. Signatories specifically call on frontier model developers to build observability and security tooling and to keep AI agent identities traceable and accountable. The phrase the signatories chose — “we have a limited window” — frames the urgency before offensive AI capability durably outpaces defense.

The private initiative follows, by three weeks, a public initiative with the same underlying logic: on August 10, 2026, California's governor announced a new state program, the “AI Cyber Defense Program,” aimed at hardening the state's critical infrastructure against emerging threats. Both texts converge on the same diagnosis — AI is widening the attack surface faster than it is strengthening defense — but take distinct institutional routes: one voluntary and contractual, the other public and budget-backed.

For critical-infrastructure sectors — energy, oil & gas, aerospace and defense — the stakes go beyond messaging. Industrial control systems (SCADA, pipelines, grid dispatch) are exactly the systems the EU AI Act classifies as “high-risk” as of August 2, 2026 when they concern critical-infrastructure safety. A voluntary cyber-coordination call, even without quantified commitments, now layers onto a landscape already shaped by a binding regulatory obligation on those same systems.

Analysis by

Cardan-AI Intelligence

Our research and analysis unit, dedicated to applied AI for business, industry and regulatory compliance.

Let's talk about your next competitive edge

A 30-minute conversation to identify your most profitable AI use cases.