What happened

Google, Microsoft and OpenAI issued a joint call Wednesday for stronger cyber defenses against AI-powered threats, according to CryptoBriefing's report. The three companies argued that defensive tooling has not kept pace with offensive capability, and that the industry needs systemic changes rather than incremental patches. The statement did not name specific incidents, but it lands after a year in which security firms tracked a sharp rise in AI-generated phishing, deepfake voice attacks on treasury desks, and automated vulnerability scanning aimed at smart contracts.

The framing was pointed. The companies said the asymmetry now favors attackers by a wide margin. Generative models cut the cost of writing convincing spear-phishing English to near zero, and code-generation tools let low-skill actors probe contracts and RPC endpoints at scale. The defenders' side, by contrast, still relies heavily on signature-based tooling and human review queues that don't scale with the volume.

Why it matters

Crypto sits at the sharp end of this problem. Exchanges and custodians hold hot wallets that clear seven and eight figures a day. DeFi protocols expose their entire attack surface publicly, by design. The economic incentive to breach a single treasury multisig or drain a lending pool dwarfs the payoff from a typical enterprise ransomware hit, and the payment rails are already crypto-native.

That's the uncomfortable read. A joint statement from Google, Microsoft and OpenAI doesn't move policy on its own, but it does mark the moment the biggest AI vendors publicly conceded their own tooling has changed the threat model. Regulators tend to follow that signal. Insurance underwriters follow it faster. Crypto firms that couldn't get cyber coverage in 2024 without an SOC 2 report should expect the audit bar to include AI red-teaming inside twelve months.