On Tuesday, the world found out about two massive security vulnerabilities in PHP, a programming language that is used by roughly 70 percent of websites. The flaw was discovered by cybersecurity researchers using a frontier model from OpenAI. If they exploited it, attackers could gain full control of servers, stealing or copying anything inside or simply taking sites offline.

Executives are twice as likely as rank-and-file workers to trust AI over their own judgment, per a survey out Wednesday morning.

It’s a worrying development. Leaders are hired for their experience and knowledge.

Outsourcing that to AI raises the risk of making decisions that don’t line up with reality, which then reverberate across companies and beyond.