The U.S. Government Blocked the Anthropic Model. It Wasn't About 'Jailbreaking' at All.

marsbitОпубліковано о 2026-06-16Востаннє оновлено о 2026-06-16

Анотація

Last Friday, the U.S. Commerce Department issued an enforcement letter that forced Anthropic to take its two most advanced AI models, Fable 5 and Mythos 5, offline. The stated reason was unspecified national security concerns, initially linked to potential "jailbreaks" of the models' safeguards. However, new details suggest the action stemmed more from a deteriorating relationship between the Trump administration and Anthropic, rather than a genuine technical threat. According to reports, the government cited a little-known export control regulation, compelling Anthropic to block access for all non-U.S. persons, including its own international employees. The company complied, shutting down the models without a court order or specific technical details from the government. Cybersecurity expert Katie Moussouris revealed she was privately shown a research paper detailing a potential safeguard bypass in Fable 5. She argued the described method was minor and did not warrant an export ban, stating that attempts to "fix" it would only weaken the model's defensive capabilities. Moussouris and other experts have since called for the order to be revoked, warning it dangerously removes advanced cybersecurity tools from U.S. defenders. Analysts like Justin Hendrix suggest the move appears retaliatory and sets a dangerous precedent, signaling that the U.S. government can unilaterally shut down a tech company's products. The incident has raised concerns about the reliability of American...

Author:Zack Whittaker

Translation: Shenchao TechFlow

Shenchao Insight: Last Friday, a letter from the U.S. Department of Commerce forced Anthropic to take down its two most powerful models, Fable 5 and Mythos 5. The government cited national security, leading outsiders to believe it was about bypassing model safeguards. However, mounting details point to another explanation: this looks more like retaliation following a breakdown in relations between the Trump administration and Anthropic. That a tech company could have its products shut down by a government order without court approval sends a signal to the entire U.S. tech industry.

The enforcement letter sent by the U.S. government to Anthropic, forcing the company to take its latest AI models offline on the eve of the weekend, should serve as a wake-up call for any U.S. tech company, not just AI labs.

First, let's review the timeline of events over the past few days. On Friday afternoon, the U.S. Department of Commerce sent a letter to Anthropic, invoking a little-known export control regulation to prohibit non-U.S. persons (including Anthropic's own employees) from accessing Fable 5 and Mythos 5, citing unspecified national security concerns. Anthropic stated it believed the letter was related to a bypass of the models' safeguards but was uncertain because the letter provided no specific details. The letter has not been made public to this day.

Anthropic's response was to shut down both top-tier models for all customers to ensure compliance. The result is that the U.S. government successfully forced a tech company to take its models offline with a swift, unilateral action that didn't even appear to require court approval.

This intervention by the Trump administration demonstrates that the AI industry is not immune to government action. For the broader tech industry, it's also a warning: comply, or we can shut you and your products down.

Axios, citing sources, described the tense situation between these two major players over the weekend, stating that what truly triggered this export order was a "personality clash" between Anthropic and the Trump administration, rather than any technical issue with the AI products themselves.

New details that emerged over the weekend further undermined the government's already shaky justification.

Veteran cybersecurity researcher and Luta Security founder Katie Moussouris wrote in a blog post that Anthropic recently showed her a private paper authored by several security researchers, describing a so-called safeguard bypass in Fable 5. (The Wall Street Journal reported the paper's authors are security researchers from Amazon.) Moussouris said Anthropic approached her to get her opinion on the paper.

Moussouris outlined in her blog how the researchers triggered this bypass but stated the bypass itself "should not have triggered export controls." The distinction is subtle: prompting the AI to "check code for security issues" versus prompting it to "fix this code"—slightly different phrasing leading to essentially the same outcome.

"The behavior described in the paper cannot truly be fixed; any attempt would only weaken the model's defensive capabilities," Moussouris said. She criticized the export control order as hasty, heavy-handed, and misguided.

Subsequently, Moussouris and dozens of top security researchers and experts called on the Trump administration to revoke the export control order, stating that removing advanced cybersecurity capabilities from U.S. cyber defenders is "dangerous."

Successive administrations have made sweeping decisions based on knowledge gaps. For example, when the U.S. government revised export laws in the 2010s aiming to control cybersecurity tools usable for both defense and offense, the wording was so broad it inadvertently risked criminalizing legitimate security and vulnerability research.

But the Trump administration's order looks like retaliation.

Justin Hendrix, editor of Tech Policy Press, said the Trump administration's move "could well raise alarms among foreign governments about the reliability of U.S. AI for critical applications." The message sent is that U.S. AI companies cannot operate free from U.S. government interference.

The Trump administration has not confirmed why it invoked this export control order. Did officials misread the report and panic? Did Amazon CEO Andy Jassy, out of caution or personal grievance, say something to senior government officials that triggered this reaction? Was it a translation error, or was this itself a tactic to pressure Anthropic—given the already tense relations? It's also possible the White House didn't anticipate the ripple effects of the letter's demands, and officials are now scrambling to clean up a mess of their own making.

In Hendrix's words, "The atmosphere is now one of suspicion, with senior officials seemingly picking favorites based on personal and political factors." The consequence is that the government has set a dangerous precedent regarding "how much control it intends to assert over the release of U.S.-made software."

This time, the government targeted Anthropic. Tomorrow, it could be any other company.

Пов'язані питання

QWhat was the official reason given by the U.S. government for forcing Anthropic to take down its Fable 5 and Mythos 5 models?

AThe official reason cited in the U.S. Commerce Department's enforcement letter was unspecified national security concerns, referenced under a little-known export control regulation.

QAccording to the article, what is the more likely real reason behind the U.S. government's action against Anthropic?

AAccording to the article, the more likely reason is a deteriorating relationship and a 'personality clash' between the Trump administration and Anthropic, suggesting the action was an act of retaliation rather than a genuine technical or security concern.

QWhat was the core argument made by cybersecurity expert Katie Moussouris against the export control order?

AKatie Moussouris argued that the 'jailbreak' or prompt vulnerability described in the research paper shown to her by Anthropic was minor and did not justify an export control order. She stated that any attempt to 'fix' it would only weaken the model's defensive capabilities.

QWhat broader warning does the article issue to the U.S. tech industry based on this event?

AThe article warns that this event sets a dangerous precedent, showing the government can unilaterally and swiftly shut down a tech company's products without court approval. It signals that no tech company, not just in AI, is immune from such government intervention.

QWhat potential international consequence of this government action is highlighted in the article?

AThe article suggests this action could make foreign governments question the reliability of U.S. AI for critical applications, as it demonstrates that American AI companies cannot operate free from U.S. government interference.

Пов'язані матеріали

Robinhood, the First Stock in the Predictive Market Concept

The online brokerage Robinhood, which previously partnered with prediction market platform Kalshi to offer event contract trading to its users, is now becoming a direct competitor. This shift began after Robinhood, through a joint venture, acquired and rebranded a CFTC-regulated exchange (now Rothera Exchange). Robinhood's motivation stems from the rapid growth of prediction markets on its platform, which significantly boosted its "other transaction revenue." Recognizing that its vast retail user base is the most critical asset, Robinhood aims to capture more value by routing orders to its own exchange instead of sharing fees with Kalshi. It strategically launched its Rothera platform during the high-traffic 2026 FIFA World Cup, successfully processing tens of millions of contracts in its initial days. This move signals a pivotal power shift in the prediction market industry: control over user distribution and access is emerging as a more decisive advantage than the underlying market infrastructure itself. The future competition may increasingly revolve around which platforms control the major user gateways.

marsbit3 хв тому

Robinhood, the First Stock in the Predictive Market Concept

marsbit3 хв тому

After Tokenization of Assets, How to Exit?

Title: How to Exit After Asset Tokenization? Author: Symbiotic Compiled by: Hu Tao, ChainCatcher Summary: Tokenization addresses how assets go on-chain but largely leaves the redemption question unresolved. While tokenized assets can settle instantly, the underlying redemption for assets like treasuries, private credit, or real estate can take from T+1 to 180 days. This gap hinders DeFi adoption of Real World Assets (RWAs). Three emerging models aim to provide instant exit liquidity, differing primarily in their capital structure and efficiency: 1. **Balance Sheet Model (e.g., Grove Basin):** A single entity (like Sky) provides immediate liquidity from its balance sheet, acting as a bridge during the settlement period. It offers simplicity and deep initial liquidity but is constrained by a single entity's capacity and risk appetite. 2. **Asset-Specific Vault Model (e.g., Upshift Clear):** Independent liquidity providers fund dedicated vaults for each supported asset, earning fees. It decentralizes capital sources but isolates liquidity and capital per asset, leading to potential fragmentation. 3. **Shared Liquidity Layer Model (e.g., Symbiotic Liquid Lane):** A shared capital pool supports multiple RWA types simultaneously. Funds remain productive between redemptions (e.g., earning yield in lending markets). Exits are settled via a competitive RFQ market. This model aims for higher capital efficiency, scalability across assets, and serves longer-duration assets like private credit. Key differentiators are: 1) Source of capital and risk bearer, 2) Redemption pricing mechanism, 3) Capital efficiency, 4) Scalability to new asset types, and 5) Composability. The shared liquidity layer model represents a move from piecemeal solutions toward scalable infrastructure, enabling T+0 exits by pooling capital, maintaining yield, and using competitive pricing, thus enhancing RWA utility in DeFi.

marsbit11 хв тому

After Tokenization of Assets, How to Exit?

marsbit11 хв тому

After Tokenizing Assets, How to Exit?

After tokenization, a key unresolved issue is providing holders with a reliable exit mechanism, as underlying asset settlement (taking days to months) lags far behind on-chain token settlement. Three primary models for instant liquidity have emerged, differing in their capital structure and efficiency: 1. **Balance Sheet Model (e.g., Grove Basin):** A single, well-capitalized entity (like Sky) provides immediate liquidity from its own reserves. This offers simplicity and deep initial liquidity but is constrained by that single balance sheet's capacity and risk appetite, limiting scalability. 2. **Dedicated Vault Model (e.g., Upshift Clear):** Independent liquidity providers (LPs) fund separate vaults for each supported asset. This decentralizes capital sources but isolates liquidity and capital, which becomes inefficient as the number of tokenized assets grows. 3. **Shared Liquidity Layer Model (Symbiotic Liquid Lane):** Independent capital providers fund shared vaults that can support multiple tokenized assets simultaneously. Capital remains productive between redemptions (e.g., earning yield in DeFi markets). Exits are settled via a competitive RFQ market where market makers bid. The article argues that the shared layer model offers superior capital efficiency and scalability. It transforms exit liquidity from an asset-specific patch into shared market infrastructure, allowing liquidity capacity to grow with overall market participation rather than being fragmented per asset. This is particularly valuable for longer-duration assets like private credit, where reliable T+0 exits can significantly enhance their utility in DeFi.

链捕手25 хв тому

链捕手25 хв тому

Anthropic's Triple Moment: Code Leak, Government Confrontation, and Weaponization

This article analyzes Anthropic's recent conflicts and strategic moves following the U.S. government's emergency halt of its new Fable model, citing national security concerns over potential "jailbreaks." The author argues this incident reveals deeper tensions between AI labs, governments, and the software industry. While critics view Anthropic's safety-focused rhetoric as marketing fear, the author suggests it serves as a commercial moat masking the company's core economic imperative: moving closer to end-users and their valuable data to avoid being commoditized. The piece outlines a coming clash between frontier AI labs like Anthropic and established software companies. Labs need real-world usage data for model improvement via reinforcement learning, creating a cycle where better products attract more users and more data. This threatens software firms who, as Microsoft's Satya Nadella warns, risk having their value captured by a few dominant models. Anthropic's controversial policy changes—initially secretly degrading Fable's performance for LLM development and expanding data retention—are framed as assertions of control, justified by its safety narrative. The company's foundational belief that it alone is sufficiently concerned about superintelligent AI dangers legitimizes its actions, from resisting government demands to shaping usage policies. The author concludes that this alignment of mission, talent, and business strategy is powerful but concerning, as it concentrates immense potential power in the hands of those convinced of their own righteous understanding.

marsbit35 хв тому

Anthropic's Triple Moment: Code Leak, Government Confrontation, and Weaponization

marsbit35 хв тому

SPCX to Surge to $400? Can Options Propel SpaceX to Become the World's Most Valuable Company?

Summary: ZeroHedge posits that SPCX (SpaceX) options trading could trigger a gamma squeeze, potentially driving the share price to an extreme $400. This scenario is based on SPCX's unique post-IPO market structure: a low float, concentrated retail buying, and the imminent introduction of options. While the $400 price is speculative, the analysis highlights the potential for outsized volatility. The core mechanism involves heavy retail buying of out-of-the-money call options, forcing market makers to hedge by purchasing the stock, creating a potential feedback loop that accelerates price gains. Vanda data shows intense retail focus on SPCX specifically, not a broad market risk-on shift. Crucially, this potential volatility stems from short-term trading dynamics and structural factors, distinct from fundamental valuation drivers like those seen in NVDA. Whether the extreme $400 scenario materializes depends on actual options market data post-launch, including open interest in call options, implied volatility, and continued buying support in the underlying stock. The article cautions that while the conditions for heightened volatility are present, the predicted gamma squeeze is not guaranteed.

marsbit46 хв тому

SPCX to Surge to $400? Can Options Propel SpaceX to Become the World's Most Valuable Company?

marsbit46 хв тому

Торгівля

Спот

Ф'ючерси