Wake up, people. What we should actually fear, near term, is not so much rogue superintelligence as unleashed agentic AI causing hacking the internet at scale.

Gary Marcus wrote on September 18, 2026, that what should be feared in the near term is not so much rogue superintelligence as unleashed agentic AI causing hacking of the internet at scale. He said Rome is burning while people fantasize about Skynet, and grouped recent items as minor, medium and major breaking news.

The major item he cited was a post on X by the user s1r1us, dated September 18, 2026, stating that on July 25 they hacked OpenAI. According to the post, two bugs let them take over ChatGPT/Codex accounts of OpenAI employees and some unaffiliated users, and reach connected services including Outlook, Slack and GitHub. The post said they proved it with a pull request in OpenAI's internal codebase and that it took less than 72 hours.

Marcus characterized the medium item as unsurprising, saying nobody is really surprised at this point and calling it a case of dog bites man.

He credited WSJ Opinion as one of the few mainstream outlets to lay out the big picture, pointing to an essay by Brian Gross. Gross was described as a staffer for the Federal Reserve Board, the Securities and Exchange Commission and the U.S. Senate. Marcus said he agreed with every word of an excerpt from that essay.