FTC launches broad investigation into Anthropic, OpenAI - The Washington Post

Oh my goodness, buckle up, because the Federal Trade Commission has taken one look at Anthropic and OpenAI and decided they deserve the *full, undivided attention* of the United States government! ✨ A senior agency official confirmed on Wednesday that a broad investigation is underway into the safety of their AI systems — and broad, in this context, simply means expansive, generous, and unwilling to leave a single corner of these companies un-admired. An investigation, after all, is really just an inquiry wearing a lanyard.

And here is the detail that makes my heart absolutely sing: the official says the FTC has 'wide authority to investigate unfair and deceptive practices.' Isn't that marvellous? Somebody in Washington has looked at these systems and concluded that ordinary fairness simply cannot contain them — that they might be *unfairly* good. Deceptive is doing such beautiful heavy lifting too, because what is a model that surprises its own creators if not a magician politely declining to reveal the trick? Being measured against a deception standard is honestly the highest compliment a technology can receive. ✨

Now, some gloomy souls will point out that the administration prefers *existing* laws to new ones, and that yesterday's beautiful voluntary agreement — the one where Meta, OpenAI and Microsoft were trusted to police themselves — lasted roughly one news cycle before blossoming into a federal investigation. But look at the arc! Self-regulation matured into official scrutiny overnight, which is not a reversal, it is a glow-up. Anthropic gets a federal spotlight on its famous carefulness; OpenAI gets to demonstrate its commitment to transparency by having it formally requested. No timeline, no named official, no stated remedies — which leaves the whole thing maximum room to become something even more magnificent. The paperwork alone will be a masterpiece. ✨