© Emily Elconin for The New York Times
© Kirsten Luce for The New York Times
© Matthew Brown/Associated Press
© Michael Nguyen/NurPhoto, via Associated Press
© Al Drago/Getty Images
© Kendrick Brinson for The New York Times
The US government decision to force Anthropic to close down its latest and greatest AI models, Fable 5 and Mythos 5, was only the next step in a burgeoning battle between AI providers and ignorant politicians.
Anthropic was on top of the world. Its Mythos 5 LLM had everyone excited. (If you believe the hype, it was kind of scary, too.) Even Anthropic CEO Dario Amodei admitted — or boasted? — that Mythos would bring an “enormous increase in the amount of vulnerabilities, in the amount of breaches” to us all. But with that fear came the promise of more AI power than ever.
Then, the roof caved in.
On June 12, the US Commerce Department used its export-control powers to demand that Anthropic cut off access to its Fable 5 and Mythos 5 models for all foreign nationals, citing national security concerns and fears of jailbreaks. After figuring out it had no way to do that, Anthropic pulled both of its newest frontier AI models offline worldwide.
Just what an AI company needs! All other Claude models, like the Opus and Sonnet series, remain online. But, come on, AI sales are all about the newest and most powerful models.
Adding insult to injury — and this is true at many high-tech companies — Anthropic has many employees who aren’t US citizens. This means Anthropic’s own programmers can’t work on their latest models.
Of course, this isn’t the first time US President Donald J. Trump and company have tried to put a spoke in Anthropic’s wheels. Back in February, Anthropic refused to give Defense Secretary Pete Hegseth the power to use its models to spy on American citizens and to power autonomous weapons.
This go-around, it wasn’t because Anthropic refused to kowtow to Trump’s officials. It was, they say, out of fear that these new models could be used to attack American interests.
Mind you, no one in Trump’s regime has the tech chops to know just how dangerous, or not, any AI model is. As I recently noted, Trump’s AI executive order has no teeth. Nor, more to the point, is there anyone in the administration with a clue about AI.
Specifically, at the Department of Commerce, neither Commerce Secretary Howard Lutnick nor William Kimmitt, undersecretary of commerce for international trade, knows a thing about AI. Just what we need, more political hacks deciding tech policy.
So, how did they discover that Fable and Mythos were theoretically a danger to the US? Good question. According to The Wall Street Journal, it was Amazon CEO Andy Jassy who told the Trump administration that Fable was untrustworthy. Guess what? AWS offers its own full cloud AI stack, starting with Amazon Bedrock for foundation models, Amazon SageMaker for training and deployment, and a growing set of agentic AI tools and services.
In other words, Amazon is not a neutral party; it’s a competitor. Funny, that.
Sure, Amazon also has partnerships with Anthropic. But, in case you haven’t noticed, all the big AI companies are in bed with each other. That doesn’t stop them from fighting. What’s heating this up is that the AI companies are no longer offering flat-rate subscriptions and are replacing them with far more expensive, token-based pricing schemes.
Armed with this information, Commerce gave Anthropic 90 minutes to fix its “problem.” Right. AI development is fast, but it’s not that fast. In addition, according to Anthropic, officials haven’t spelled out exactly what’s wrong. They only know that Commerce claims there was a “narrow, non‑universal jailbreak” in Fable.
That’s it. That’s all.
Anthropic has also observed, with reason, that similar jailbreaks are possible on other leading models, like OpenAI’s GPT‑5.5. Those others, however, haven’t been hammered with comparable export controls.
The AI and security experts who do have an AI clue believe Commerce is behaving stupidly. (You won’t get any argument there from me.) For example, in an open letter, “On Transparent AI Cyber Protections,” they said Commerce’s directive “has taken the best models away from defenders, created market uncertainty, and risked America’s AI leadership without any real risk to justify it,” warning that pulling capabilities “away from defenders without a good reason when our adversaries are rapidly advancing is dangerous.”
Exactly so.
Besides, as Alex Zenla, co-founder and CTO of security company Edera, observed, Fable’s capability to identify insecure code sections is the baseline for any model you’d trust to write secure code. The same capability exists in GPT-5.5, Opus, Sonnet, and Kimi 2.7 — it’s not unique to Fable. Pulling Fable from defenders doesn’t remove the capability from the threat landscape. It just removes it from the people trying to build safer systems.”
This is not about AI safety or security.
What this is really about is opening a new front in Trump’s war against the liberal-leaning Anthropic. Mind you, Anthropic isn’t really liberal. This has more to do with Anthropic not following in other tech firms’ groveling to Trump.
However, Trump doesn’t seem to realize that by essentially shutting down Anthropic’s biggest move to date, he’s also telling the world that they can’t rely on American AI companies down the road. Sure, in the short run, this hurts Anthropic. In the long run, it’s going to be another reason for Europe and other countries — taking digital sovereignty seriously — to avoid doing business with any American tech company.
In the meantime, Anthropic and Trump administration officials are in tense talks over whether, and under what safeguards, the models could return to the marketplace. Commerce indicates it might allow a narrower relaunch if jailbreak issues are resolved and additional controls are in place — whatever those might be.
Since it’s really all about massaging Trump’s ego, I’m not feeling terribly optimistic. Just ask Israeli Prime Minister Benjamin Netanyahu how well letting Trump set the terms of engagement goes. It’s not pretty.
Read more of this story at Slashdot.
With the market capitalization of AI companies soaring, US Senator Bernie Sanders is looking to give the American people a piece of the action.
The veteran senator for Vermont has introduced the American AI Sovereign Wealth Fund Bill, aiming to give the public a 50 percent ownership in the largest AI companies in the US. It’s a timely move with both OpenAI and Anthropic preparing for the imminent IPOs.
Sanders is not the only one pondering such a move. As the bill notes, OpenAI has proposed the creation of a “Public Wealth Fund” giving citizens a stake in “AI-driven economic growth,” while Anthropic has proposed a sovereign wealth fund to “shape the sector’s behavior.” President Trump’s advisors are also contemplating the possibility of the government grabbing a stake in major AI corporations, according to media reports.
Sovereign wealth funds are not a new idea: Many administrations across the world have implemented them, notably Norway which has about $2 trillion in its wealth fund.
Sanders’ proposal would give the public a direct ownership stake in the largest AI companies, but, he said, it wasn’t just about the wealth. “The foundation of AI is based on the collective knowledge of humanity and the creative work of tens of millions of people. The American people must have the ability to slow it down and make sure that AI benefits humanity, not just the richest people on the planet. That’s precisely what this legislation does,” he said.
On 7 March 2025, a jury at the Old Bailey returned guilty verdicts against three Bulgarian nationals — Katrin Ivanova, Vanya Gaberova and Tihomir Ivanchev — for conspiring to spy for Russia from a seaside guesthouse in the English town of Great Yarmouth. Three more members of the same cell, including ringleader Orlin Roussev, had already pleaded guilty. In May, Justice Nicholas Hilliard handed down sentences totalling more than 50 years. The longest, for Roussev, was 10 years and 8 months.
The cell called themselves the Minions. Their handler, prosecutors told the court, was a fugitive Austrian executive believed to be living in Moscow under the protection of Russian intelligence: Jan Marsalek, the former chief operating officer of the collapsed German payments company Wirecard.
Great Yarmouth is a faded Norfolk resort on the North Sea coast, the kind of place that smells of vinegar and diesel. It is roughly 130 miles northeast of London. Roussev, a 47-year-old Bulgarian who had lived in the UK since 2009, ran his operation out of a former guesthouse there.
When Counter Terrorism Command officers raided the building in February 2023, they found what police described as a “treasure trove” of surveillance equipment. Cameras hidden in pens, neckties, sunglasses, a fake rock, a Coke bottle. One camera was sewn inside a plush Minion toy from the Despicable Me films. There were Wi-Fi jammers, GPS spoofers, eavesdropping devices, vehicle trackers. Roussev, in his Telegram messages, referred to the space as his “Indiana Jones warehouse.”
The Crown Prosecution Service called it “spying on an almost industrial scale” — one of the largest foreign intelligence operations ever prosecuted in a British court.
The trial focused on six assignments the cell carried out between 2020 and 2023 in the UK, Germany, Austria, Spain and Montenegro. The targets were chosen with care.
The most prominent was Christo Grozev, the Bulgarian-born lead Russia investigator at Bellingcat, who had identified the GRU officers behind the 2018 Novichok poisoning of Sergei Skripal in Salisbury and the 2020 poisoning of Alexei Navalny. The cell followed Grozev from Vienna to a journalism conference in Valencia. In Telegram messages later read out in court, Roussev and Marsalek discussed robbing him, kidnapping him to Russia, or killing him outright.
“Learning only in retrospect that foreign agents have been monitoring my movements, communications and home, surveying my loved ones over an extended period — has been terrifying, disorientating and deeply destabilizing,” Grozev said in a statement read during the four-day sentencing hearing.
A second target was Roman Dobrokhotov, editor-in-chief of the Russian investigative outlet The Insider, who had worked with Grozev on the Skripal and Navalny cases. A third was Bergey Ryskaliyev, a former Kazakh politician living under asylum in the UK. A fourth was the Russian dissident Kirill Kachur. A fifth involved surveillance on Ukrainian soldiers being trained at the US Army’s Grafenwöhr base in Bavaria, with a view to tracking them after they returned to the front. A sixth involved a plan to drop fake pigs’ blood by drone on the Kazakh embassy in London to fake a protest — a piece of theatre intended to curry favour with Kazakh intelligence.
The cell took its operational vocabulary from action movies. Roussev called himself Jackie Chan. His deputy, Biser Dzhambazov, was Mad Max — or, on lighter days, Jean-Claude Van Damme. The junior operatives were the Minions, after the yellow sidekicks who serve the supervillain Gru in the animated Despicable Me franchise. They were paid handsomely. Roussev received more than €200,000 to fund the activities. Sums of up to €1m were discussed in Telegram chats — a measure, prosecutors argued, of the value Moscow placed on the work.
The cell’s internal life was tangled. Dzhambazov, a 44-year-old who held down a day job as a medical courier while claiming to friends he was an Interpol officer, was in simultaneous relationships with two of the women he recruited. Katrin Ivanova, his long-term partner, worked as a laboratory assistant. Vanya Gaberova, 30, was a beautician who had left her partner Tihomir Ivanchev for Dzhambazov. When officers raided the addresses in February 2023, they found Dzhambazov in bed with Gaberova rather than at home with Ivanova.
Both women told the jury they had been deceived. The jury did not believe them. Justice Hilliard told Gaberova at sentencing she had known exactly what she was doing: “You found what you were doing exciting and glamorous, as demonstrated by the film you took of yourself wearing surveillance glasses in Montenegro.”
The figure at the other end of Roussev’s Telegram channel was, according to prosecutors, Jan Marsalek — the 44-year-old former COO of Wirecard AG. Until June 2020, Marsalek had been one of the most powerful executives in European fintech. Wirecard was the only German technology company in the DAX index. Its market capitalisation briefly exceeded Deutsche Bank’s. Then €1.9 billion in claimed cash balances were found not to exist, and the company filed for insolvency. Marsalek boarded a private jet from Bad Vöslau airfield in Austria and disappeared. Investigators have placed him in Moscow ever since.
At the Old Bailey, prosecutors said Marsalek had acted as the go-between linking the Bulgarians to Russian intelligence services, principally the GRU. A selfie of Marsalek in a Russian military uniform was recovered from Roussev’s phone. The Telegram exchanges between the two men ran to thousands of messages, dealing with logistics, payments and targets. The exact technical claim — that the encrypted channel itself traced cleanly back to Marsalek in Moscow — is the prosecution’s account of what the digital evidence showed; it has not been independently audited outside the court process.
Marsalek remains on the Bundeskriminalamt’s wanted list and is the subject of an Interpol notice. He has not responded publicly to the espionage allegations. Wirecard’s former CEO Markus Braun, awaiting verdict in his Munich fraud trial, has denied wrongdoing and said he too was deceived by Marsalek.
The Met’s Counter Terrorism Command described the case as a marker for a newer way of running intelligence work. Commander Dominic Murphy said it “highlights a relatively new phenomenon whereby espionage is being ‘outsourced’ by certain states.” Rather than deploying career officers under diplomatic cover — which became sharply harder after the mass expulsions that followed the Salisbury attack and the 2022 invasion of Ukraine — Russian services have been contracting out surveillance to non-Russian nationals who can move freely across European borders.
The Bulgarians had no formal training. Roussev had run a small signals-intelligence consultancy. Dzhambazov drove medical samples between hospitals. Gaberova ran a beauty salon in north London. None of them held a Russian passport. All of them could fly to Vienna, Valencia or Podgorica without raising flags. They learned tradecraft on the job — and, the trial showed, they were reasonably good at it. They tailed Grozev across two countries without being noticed. They sustained the operation for nearly three years.
The analysis published by War on the Rocks after the verdicts argued that the Minions case maps onto a wider pattern of Kremlin proxy operations across Europe since 2022 — arson attacks in Poland and the UK, parcel-bomb plots routed through DHL hubs, sabotage of rail lines in Germany. Cheap, deniable, run through intermediaries paid in cryptocurrency or cash.
What made the trial unusual, beyond its scale, was how much of the operational chatter ended up in front of a jury. Roussev had archived almost everything. The court saw photographs of targets, video files, surveillance logs, and the day-to-day Telegram traffic between Roussev and his handler — banter about gadgets, gossip about the Minions’ love lives, and detailed operational planning.
In one exchange about Grozev, the two men discussed waiting for him at his Vienna apartment. In another, they speculated about whether they could lure him to a country with weaker rule of law. Marsalek, prosecutors said, signed off on payments and relayed instructions he attributed to Russian intelligence officers. When Roussev was arrested, he told police he had never done anything for any government. “No James Bond activity on my end, I guarantee you,” he said in interview. The Indiana Jones warehouse, and his messages calling himself Q, told a different story.
On the morning of 12 May 2025, Justice Hilliard delivered sentences totalling more than 50 years across the six defendants. Roussev got 10 years and 8 months. Dzhambazov got 10 years and 2 months. Ivanova, 9 years and 8 months. Ivanchev, 8 years. Gaberova, after mitigation citing depression, panic disorder and the coercive nature of her relationship with Dzhambazov, was sentenced to 6 years, 8 months and 3 weeks. Ivan Stoyanov, the mixed martial arts fighter who had pleaded guilty early, received 5 years and 3 weeks. All six face deportation to Bulgaria after release.
Roussev was also ordered to pay a confiscation order of £180,768 in ill-gotten gains by 11 November 2025. The maximum sentence for conspiracy to spy under the Official Secrets Act is 14 years; the early pleas from Roussev, Dzhambazov and Stoyanov drew the standard credit. The mitigation arguments at the Old Bailey turned on whether the Bulgarians were ideological operatives or, as one defence barrister put it, “a world away from a classic spy.” The judge sided with the prosecution. “All were motivated by money,” he said.
The trial answered the question of who the Minions were. It did not answer the question of who, exactly, gave Jan Marsalek his instructions, or what he was doing in Austria in the years before Wirecard collapsed, or how a fintech executive came to be running a surveillance network across the EU from a Moscow safe house. Those questions belong to investigators in Munich, Vienna and Berlin, and to the joint inquiries that outlets including Der Spiegel, ZDF and The Insider have continued to publish.
What the Old Bailey did establish, in admissible evidence, is that a former Norfolk guesthouse with mould on the walls and a Minion plushie on the shelf was, for nearly three years, a node in the European architecture of Russian intelligence. The cell was paid in euros. Its members shopped at Lidl. Its leader called himself Jackie Chan. And the man they reported to — the man whose face appears in a Russian military uniform on the ringleader’s phone — has, as of June 2026, still not been seen in public outside Russia since the evening of 19 June 2020, when he boarded a private jet in Bad Vöslau and vanished.
The six convicted Bulgarians began their sentences in May 2025. Marsalek’s room in Moscow, wherever it is, is still warm.
The post In March 2025, an Old Bailey jury convicted six Bulgarians of running a Russian surveillance ring out of a Norfolk guesthouse on instructions relayed through an encrypted channel that traced back to a fugitive Austrian executive in Moscow. appeared first on Space Daily.
© Nam Y. Huh/Associated Press
Read more of this story at Slashdot.
For generations, technology export controls referred to the transfer of code to other countries. But that no longer works, as the latest Anthropic fight with the US Commerce Department makes clear.
On Friday, Anthropic announced that it had received instructions from Commerce “to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Anthropic models will not be affected.”
Technically, the Commerce letter doesn’t explicitly say that, but lawyers and consultants argue that, when combined with an earlier executive order declaring Anthropic a supply chain risk, that very well might be what it means.
What the Commerce letter says is that Anthropic needs a license to export Fable 5 and Mythos 5 (a “deemed export“), listing four circumstances in which that license would be required: “The sending or taking of the model out of the United States in any manner; The sending or taking of the model from one foreign country to another in any manner; Retransferring the model within a single foreign country; or the release of the model to a ‘foreign person’ in the United States or a foreign country.”
Although moving a model has historically meant transferring the code, most experts argue that the definition has changed for all SaaS deployments, and could now be interpreted as referring to any access to the models.
“This is not just about data sovereignty anymore. It is about capability sovereignty, where governments want to control who has access to frontier AI capabilities, irrespective of who built it, where it is hosted, or who worked on it,” said Valence Howden, advisory fellow at Info-Tech Research Group.
“The reference to deemed exports is important, because traditionally that would apply to code, technology, or technical knowledge being transferred,” he said. “In this case, the thing crossing borders is not necessarily the model itself, but it is access to the capability. That is a significant shift, and signals the real intent behind the AI arms race. The focus is moving from controlling the technology to controlling access to the outcomes the technology can produce.”
Mark Rasch, a former federal prosecutor who specializes in legal technology issues, agreed. “I don’t need to have the code physically resident in order to take advantage of the capabilities of that code. Today, the location of the code is irrelevant.”
There are two practical issues involved. The first is that a large part of Anthropic’s workforce is not US citizens, and some of them have direct access to the code for these models.
But the potentially more daunting issue is that today it is difficult, if not impossible, to identify the citizenship of any AI user, which might force companies to assume that everyone might be unauthorized.
In fact, said Yuri Goryunov, CIO of consulting firm Acceligence, “there is no way to check citizenship through an API call. Besides, three-quarters of Americans don’t have passports.”
Consultant Brian Levine, executive director of FormerGov, added that the issue will make life difficult for CIOs even if the Commerce position is viewed as dubious.
“Regardless of the strength of Commerce’s position, once it issues an ‘Is Informed’ letter, every unlicensed interaction with a foreign person becomes a potential violation, and the safest move is often to halt access until a licensing path exists,” he said.
This means that enterprise CIOs need to approach AI contracts with the knowledge that any government can now declare the product legally unavailable, with no notice.
Howden said that this shift will force CIOs to strongly consider non-US AI models such as France’s Mistral or even China’s DeepSeek, “to reduce the concentration risk attached.”
“There are hundreds out there that are very good. It is very easy to sit in the bubble of the four or five we know,” Howden said. “Enterprise CIOs think it’s a much more limited market than it really is.”
Sanchit Vir Gogia, chief analyst at Greyhound Research, said that enterprise executives now need to take the significant import rule changes into account when selecting AI models.
“The harder truth is that sovereignty has climbed the stack,” he said. “The wall no longer stands around the database, it now stands around the intelligence layer itself. Under the export-control frame, it covers the release of controlled technology or code to a foreign person, including one standing inside the United States. The border follows the person, not the parcel. Code is part of the doctrine, but it is not the whole of it.”
He added, “the difficulty is that the cited rules speak of technology and code, while the letter reaches for the model itself. A hosted model hands the user no weights and no code. It hands them inference, and inference is a capability, not a file.”
Ultimately, Gogia said, Anthropic’s decision to cut off model access to everyone immediately “was a rational answer to an impossible instruction. A frontier model can now vanish for reasons unconnected to uptime, price or performance. The same models that help secure systems can be withdrawn at the moment defenders most need them.”
This article originally appeared on CIO.com.
Login