Two weeks into the ban that caused Anthropic to pull its powerful cybersecurity-oriented models, Mythos 5 and Fable 5, from the market, the Trump administration is softening its stance.

It is now allowing Anthropic to make Mythos 5 available to more than 100 specific U.S. government agencies and companies, including allowing the non-American employees at those organizations to access to the model, both Semafor and Reuters report. This list also includes Anthropic’s own non-American employees, who were included in the original ban that forbade non-Americans from accessing the models.

“I have determined that appropriate safeguards are in place to permit certain trusted partners to access the Claude Mythos 5 Model,” Commerce Secretary Howard Lutnick wrote to Anthropic’s chief compute officer Tom Brown on Friday, according to the missive seen by Semafor.

Apparently, the administration did not address the release of Fable 5 in this directive. This is a version of Mythos 5 that was widely released a couple of days before the ban because it was said to have more protections. Both models were pulled after those guardrails were allegedly bypassed easily by security researchers. Anthropic did not immediately respond to our request for comment.

Elon Musk is eyeing an acquisition of Mesh Optical Technologies, a startup founded by three former SpaceX engineers last year developing hardware for fast data center communications.

The potential acquisition, which was revealed in a Federal Trade Commission filing and first reported by Bloomberg, confirmed the agency expedited its antitrust review.

Mesh Optical came out of stealth in February when it announced that it raised a $50 million Series A led by Thrive Capital.

Before founding Mesh Optical, the startup’s co-founders, Travis Brashears, Cameron Ramos, and Serena Grown-Haeberli, developed the optical communication links that keep thousands of SpaceX’s Starlink satellites interconnected.

The Mesh co-founders saw an opportunity to develop optical transceivers for terrestrial data centers, as light-based hardware is faster and more energy-efficient than traditional electrical-based systems.

SpaceX has recently entered into agreements with Anthropic, Google, and the open-source AI developer Reflection AI to provide them with compute capacity at its data centers, generating a substantial new revenue stream for the newly public company. Acquiring Mesh could eventually allow SpaceX to improve the efficiency of its data centers, whether they are located on Earth or, in the future, in space.

Y Combinator-backed insurance tech startup Corgi became embroiled in yet another controversy earlier this week when Papermark, maker of open source data room software, accused Corgi of stealing its software and passing it off as its own.

Corgi denies this, telling TechCrunch. “No code was used from Papermark.”

But there were reasons why people believed the initial allegation, which was made by Papermark co-founder Marc Seitz on X, concerning Corgi’s newly released product called Dataroom.

Seitz’s post blew up because he shared screenshots showing Corgi’s product using the same language for the same features as Papermark’s, word for word. Deal room software is essentially secure document sharing. It is famously used by startups to pitch VCs and send them supporting materials for due diligence.

Seitz went as far as to call Corgi’s new product copyright and license-infringing, and “fraud.”

Corgi’s co-founder and CEO Nico Laqua saw the tweet and promised to investigate. Soon after, he posted on X his full denial with receipts of his own, showing that the code was different between the two products.

While he strenuously pushed back on the allegations of a license violation (“‘stole my enterprise-code’ is a different claim than ‘copied my style,” Laqua argued), he did admit that relying on a vibe-coding design led to the replica features.

“Looking back, we should’ve leaned more into our own language and visual choices instead of taking cues from existing products in the space, and that’s on us,” he posted.

A Corgi spokesperson confirmed to TechCrunch that the offending features were vibe-coded and said they have already been changed, downplaying the situation.

“The issues were isolated to visual elements on two peripheral settings pages,” the spokesperson told us, adding that these elements were “immediately updated” and that “our team confirmed that no code was used from Papermark.”

Laqua and the spokesperson also accused Papermark of making these accusations because Corgi is offering a less expensive product. “I get that this stings since we’re putting out something mostly free that competes with his SaaS. I’d be mad too,” Laqua wrote of Seitz. (Seitz has not yet responded to our request for comment.)

Yet, this clearly wasn’t just sour grapes when identical features and wording were used.

It’s about a new question: If vibe coding makes it so easy to copy the look, feel, and every function of another’s work, while not copying every line of the code itself, how much does it matter if the source isn’t identical?

Obviously, legally speaking, it’s the only thing that matters. So this is not the same as the controversy over Y Combinator alum PearAI, a 2024 startup that admitted to cloning another open source project and releasing it under its own license.

Morally speaking, this is ambiguous and will become increasingly common.

As fellow YC alum and founder of the agent operating system OpenProse, Dan Barrett explained on X: “In a world where a bot can trivially copy 1:1 the structure of something even if the character-level code diverges … what makes one unacceptable and the other not? existing IP law, incidental to the old world? is there not some greater principle at work here?”

Corgi is now vigorously trying to clean up any reputational damage. It has issued a cease and desist letter to Seitz demanding that he take down the tweet, the company confirmed to TechCrunch.

The founder of Hello World Cafe, which somewhat competes with Corgi’s coffee shop business, also contends on X that he got a cease and letter from Corgi’s lawyers for his tweet that joked about the Dataroom controversy. Though X still remembers. There have been hundreds of comments and countless subtweets. (Corgi also offers a 24-hour coffee shop, with plans to open more, Laqua recently said on Harry Stebbings’ podcast.)

This latest hullabaloo adds to a growing list of chatter around Corgi. The two-year-old startup, for instance, has a growing reputation for being litigious. It’s already sued various former employees.

Laqua also recently went viral for his comments on Stebbings podcast about how he expects employees to work 7 days a week. “Whatever you can done in five days, I promise you, you’ll get more done in six and seven,” he said.

That is, of course, the fallacy of startup hustle culture. Decades of research repeatedly conclude that human productivity is not a quadratic equation. While sprints can be effective (and build camaraderie) for short-term problems like the site going down, the research shows that, as a matter of routine, more hours of work reduces productivity, not the other way around.

The startup also got tongues wagging for how fast it has raised money with increasing valuations, even by AI-startup standards. Last month, Corgi raised a $106 million Series B1 raise, valuing the company at $2.6 billion, just three weeks after announcing a $160 million Series B at a $1.3 billion valuation and four months after its $108 million Series A.