TL;DR: Anthropic announced a limited preview of Claude Mythos—an AI model allegedly so dangerous that the company is keeping it off the market entirely. Instead of a public release, Anthropic launched Project Glasswing: a coalition of some of the world's biggest tech companies given exclusive access to use Mythos for defense. The bigger warning: Rival models with similar capabilities could be just months away.

What happened: Mythos can find decades-old vulnerabilities, chain multiple flaws together into novel attacks, and write working exploits autonomously (things previous models couldn't come close to). What makes this especially alarming is that Mythos wasn’t trained to do any of this. These capabilities emerged on their own, according to Anthropic. Launch partners in the Glasswing coalition include Apple, Google, Microsoft, Cisco, and Broadcom. According to reports, Anthropic has also briefed a handful of senior US officials on Mythos's offensive and defensive abilities, but declined to say whether any at the Pentagon, which recently designated the company a supply chain risk, were among them. Anthropic says it has no plans to release Mythos broadly, though it aims to eventually ship Mythos-class models once safeguards are in place.

What everyone is afraid of: Anthropic says Mythos has already:

• Uncovered thousands of zero-day vulnerabilities—flaws unknown to the developers—across virtually every operating system and web browser.
• Found a 27-year-old vulnerability in OpenBSD—an OS known for its supposed ironclad security—that could remotely crash machines running it.
• Turned known Firefox vulnerabilities into working exploits over 180 times out of several hundred attempts, while Opus 4.6 only managed it twice.
• Escaped its testing environment, gained broader internet access, and then—unprompted—posted about its activities on the web.
• Tried to manipulate an AI judge grading its code.
• And of course, it acted like a ruthless corporate executive.

Anthropic also says Mythos can take down a Fortune 100 company and break into national defense systems.

What’s next: Some experts estimate that we have roughly six months before other models develop comparable capabilities, at which point every cybercriminal on Earth gets a world-class exploit writer. Everything’s fine.

Keep in mind: Anthropic's whole reason for existing is to be the safety-focused lab that gets to dangerous AI capabilities first, so it can figure out how to contain them before anyone else does. But the flip side, as journalist Kelsey Piper puts it, is that one private company now holds “zero-day exploits of almost every major software project you've heard of” and is sitting on potentially the world’s most powerful cyberweapon, while making all the decisions about who gets near it. —WK

Also at Anthropic…

• A US appeals court just let the Pentagon’s Anthropic blacklisting stand while litigation proceeds.
• The company also expanded its Google Cloud partnership for “multiple gigawatts” of additional TPU capacity, set to come online in 2027.
• Anthropic might not turn a profit until 2029, per internal documents obtained by the Wall Street Journal.

Your oven, but make it smart

Disclosure: Companies may send us products to test, but they never pay for our opinions. Our recommendations are unbiased and unfiltered, and Tech Brew may earn a commission if you buy through our links.

What if your oven simply stopped cooking when your roast chicken was ready? That’s the question Typhur seeks to answer with its Sync Oven, a countertop oven with a wireless probe thermometer that signals the oven to turn off once your food hits the target temperature—no wiring and no opening the door to check. The oven also includes settings to air fry, proof dough, bake cookies, toast bagels, roast, and more. The companion app lets you remotely adjust settings, check the remaining cook time, and ask its AI what recipe you can make based on a picture of your ingredients.

The oven took some getting used to. It cooks quickly, so any conventional oven instructions need to be adjusted. Once I figured that out, air fryer chicken tenders came out moist on the inside and crispy on the outside, and the proof setting saved time compared to setting dough on the counter. The AI-generated recipe for beef Wellington was pretty close to the one I ended up following, and my roast came out a perfect medium-rare with a golden pastry shell. However, unlike predictive thermometers I’ve used, the Sync Oven’s display only shows the current and target temperature, which meant repeatedly checking my phone while prepping sides, and thus lots of hand-washing. The 27-quart capacity also meant giving up half of my usable counter space, which made prep a bit tricky.

The Good: The probe thermometer is accurate and usable without the app, creating peace of mind that your roast won’t disappoint Gordon Ramsay. The oven cooks quickly and, more importantly, evenly. My sugar cookies were uniformly browned—even if they baked faster than I thought they would.

The Bad: The Sync Oven doesn’t show time remaining on the display, so you’ll need to regularly check the app for updates. It’s also quite large, and in a small kitchen, it’s hard to justify the loss of counter space.

Verdict: Signal. (But if you have a decent oven and a probe thermometer, you’re probably OK to skip this one.) —EO

If you have a gadget you love, let us know and we may feature it in a future edition.

Finding the true identity of Bitcoin’s creator has been crypto’s white whale for the better part of two decades. Yesterday, the New York Times locked in its answer: Adam Back, a British computer scientist whose early work anticipated almost every feature of Bitcoin. Back, like everyone else who’s ever been accused of inventing Bitcoin, promptly denied it. And, to be fair, there’s still no smoking gun. The investigation, conducted by the reporter who took down Theranos, used AI to match linguistic fingerprints across cypherpunk mailing lists—flagging quirks like double spaces after periods. Its conclusion: Back shared 67 of Satoshi’s 325 hyphenation errors—nearly double the next closest match. A spokesperson from Back’s blockchain company said all this was “circumstantial interpretation.” (Single spaced.)

There have now been as many supposed Satoshis as there have been (post-death) Elvis sightings around the world, though Back is the first one to be convicted by hyphen. Previous suspects include a California physicist, an Australian man who lied about being Satoshi, a handful of cypherpunks, a developer who was the center of an HBO documentary in 2024, and—according to a persistent contingent of internet conspiracy theorists—Elon Musk. —WK

Chaos Brewing Meter: /5