Cybersecurity and software stocks tumbled Friday as traders reacted to Anthropic's new Claude Mythos AI model, which can detect critical code vulnerabilities that human testing missed. The S&P 500 software and services index fell 1.6 percent, bringing its year-to-date decline to 26 percent. A Goldman Sachs basket of US software stocks dropped 5 percent, extending sharp losses from recent months.

Analysts noted that Middle East tensions had temporarily distracted investors from Anthropic's announcement, but attention has now returned to the potential disruption. Mike O'Rourke at Jones Trading said the Claude Mythos model has "significantly escalated AI disruption fears." Treasury Secretary Scott Bessent and Federal Reserve Chair Jay Powell earlier this week summoned major US banks to discuss cyber risks posed by the new technology.

The selling hit information security groups particularly hard. CrowdStrike fell 4 percent, Palo Alto Networks dropped 7 percent, and Datadog declined 3 percent. Private credit firms with software exposure also declined, with Ares down 4 percent and Blackstone off 2 percent. Anthropic's Claude Mythos Preview identified thousands of previously undiscovered vulnerabilities, including a 16-year-old flaw in widely used video software that automated testing tools had executed 5 million times without detection.