This week, we and others reported that OpenAI has been in early talks with investors about a new private round that could value it at $1.2 trillion or more. That follows CEO Sam Altman’s recent statement that an initial public offering wouldn’t happen this year, not that anyone was actually expecting that!

There are plenty of unanswered questions about the recent turn of events. But the biggest—why stay private?—seems obvious. It buys the company more time to rev up sales without the pressure of disclosing on how much it’s paying for that growth, such as offering margin-crushing discounts to lure business from rival Anthropic. 

While the AI industry ties itself in knots over models it thinks will pose security threats to the internet and perhaps to humanity itself, researchers are warning about flaws in AI coding agents from Anthropic, OpenAI, Google and Microsoft that pose immediate risk to businesses.

For instance, a recently discovered (and now largely fixed) software flaw in Claude Code, Codex, Gemini CLI and GitHub Copilot would have let attackers hijack other people’s coding agents without them noticing, according to new research shared exclusively with The Information.

The research, published by Sequoia Capital-backed cybersecurity startup Air, discovered the same flaws in the way all four of those coding agents handle “skills” that direct agents to refer to a set of instructions and files to execute specific tasks.