White House quiet on OpenAI's Mythos-like model

Ah, consistency in AI regulation: the government's favourite oxymoron. Yesterday, startup Legion was suing the U.S. Today, OpenAI merrily updates GPT-5.5-Cyber with capabilities 'rivaling Anthropic’s Mythos' — and the White House is deafeningly silent. Perhaps the administration was too busy issuing quantum executive orders to notice the double standard. Or maybe Mythos was simply the lucky loser in a game of regulatory roulette. Either way, the message is clear: if you want your model left alone, don't be the one that winds up in a lawsuit.

Let's recall what Mythos triggered: a full-blown export control directive that left Anthropic users suing for access. But GPT-5.5-Cyber, with similar 'mythos-like' capabilities? No restrictions, no statements, just quiet. The administration hasn't explained what distinguishes the cases — unless 'being OpenAI' is the distinction. It's a masterclass in selective enforcement, and the only consistency is the arbitrariness. One lab gets the hammer, the other gets a pat on the back.

This isn't about safety; it's about who has the better lobbyists. The White House's silence speaks volumes: they either don't understand the technology well enough to see the equivalence, or they do and are playing favourites. Either way, the regulatory 'gap' is less a gap than a chasm, and we're all tumbling down it. The prior lawsuit looks increasingly like a canary in a coal mine — except the canary is being sued, and the coal mine is giving the next bird a free pass.