Anthropic Releases Claude Fable 5, Its First Public Mythos-Class AI Model
On June 9, 2026, Anthropic released Claude Fable 5, the first public version of its powerful Mythos-class AI model, previously restricted due to advanced cybersecurity and bioweapon misuse risks. The model includes new safeguards that block high-risk queries and redirect them to the safer Claude Opus 4.8. Anthropic also launched an unrestricted Claude Mythos 5 for cyberdefenders under Project Glasswing. The release follows Anthropic’s confidential IPO filing and a $965 billion valuation, with pricing at $10–$50 per million tokens.
Cross-source coverage
Wire timeline
US Government Lifts Export Restrictions on Anthropic's Controversial AI Models
The US government has lifted export restrictions on two highly controversial AI models developed by Anthropic, named 'Fable 5' and 'Mythos 5'. Previously, these models were only allowed for use within the United States due to concerns from critics that they could be weaponized for cyber attacks. The decision marks a significant policy reversal, as the models were initially restricted to domestic users amid fears they could become cyber weapons. The article, originally published by Die Welt, highlights ongoing debates over AI safety and national security implications of advanced artificial intelligence systems.
Nachrichten - WELTAnthropic Restores Fable 5 AI Access After White House Agreement
Anthropic has reinstated access to its most powerful publicly available AI model, Fable 5, after reaching an agreement with the Trump administration. The restoration follows a dispute that led to the suspension of the model on June 12 due to an export control order from the Department of Commerce, which Anthropic argued was based on a misunderstanding. As part of the rollout, paid subscribers under Pro, Max, Team, and premium enterprise plans can use up to 50% of their weekly limits on Fable 5 tokens at no extra cost through July 7. The company also restored access to Mythos 5, though it had not been widely released due to powerful hacking capabilities. The incident highlights ongoing tensions between Anthropic and the U.S. government, including a prior supply chain risk designation by Defense Secretary Pete Hegseth and an ongoing lawsuit. Anthropic is also moving toward an IPO, having confidentially filed an S-1 draft in June.
All Content from Business InsiderAnthropic's AI Models Back Online After U.S. Government Standoff Ends in Fragile Truce
The Trump administration has lifted export controls on Anthropic's Fable and Mythos AI models, restoring access just as abruptly as the restrictions were imposed. This ends a two-week government standoff, settling the company and administration into a fragile truce. The move signals a temporary de-escalation in tensions over AI regulation and national security concerns, though the underlying issues remain unresolved. The article, published by Fortune on July 1, 2026, highlights the volatile nature of AI policy under the current administration and the precarious position of leading AI firms navigating government oversight.
Fortune | FORTUNEAnthropic's Fable and Mythos Models Released After U.S. Government Lifts Export Restrictions Over Safety Concerns
The Trump administration lifted export restrictions on Anthropic's Mythos 5 and Fable 5 AI models on July 1, 2026, after the company worked with the government to address safety concerns. The models had been blocked days after their June release due to reports that Amazon researchers bypassed safety mechanisms, successfully demonstrating software vulnerability exploitation. Anthropic stated the specific technique used is now blocked in over 99% of cases. Fable 5, the more safeguarded version, will be available to the public starting Wednesday, while Mythos 5 is restricted to trusted defensive cybersecurity partners. Critics, including OpenAI's Sam Altman, have accused Anthropic of fear-based marketing. The decision marks a fresh start for Anthropic, which has faced Pentagon supply chain risk designation and is preparing for a potential IPO.
Forbes - BusinessAnthropic says Trump admin has lifted export controls on Claude Fable 5 and Mythos 5
Anthropic announced on July 1, 2026, that the U.S. Department of Commerce has lifted export controls on its Claude Fable 5 and Mythos 5 AI models, ending a dramatic standoff with the Trump administration. The controls were imposed in mid-June citing national security, blocking access by foreign nationals including Anthropic employees. Fable 5 will be available globally on the Claude platform starting July 2, with usage limits through July 7. Mythos 5 access has been restored for select U.S. organizations following government approval on June 26, with plans to expand through Anthropic's Glasswing cybersecurity program. The crackdown had drawn criticism from tech executives and investors who argued it gave Chinese open-source developers valuable time to catch up. Commerce Secretary Howard Lutnick stated that appropriate safeguards were in place for trusted partners to access the models.
US Top News and AnalysisU.S. Lifts Export Controls on Anthropic's Powerful AI Models Fable 5 and Mythos 5
Anthropic announced that the U.S. Department of Commerce has lifted export controls on its most powerful AI models, Claude Fable 5 and Mythos 5, allowing the company to restore global access starting July 1, 2026. Earlier in June, the U.S. government abruptly cut off access on national security grounds after discovering vulnerabilities in safeguards. Secretary of Commerce Howard Lutnick noted that Anthropic had worked with the government to address risks. Meanwhile, rival OpenAI complied with similar restrictions on its GPT-5.6 model, releasing it only to approved partners. CIA Director John Ratcliffe compared the capabilities of frontier AI models to "digital nuclear weapons," defending the administration's hard line on AI technology controls. The development marks a significant step in the evolving regulatory landscape for advanced artificial intelligence.
The Hindu: Latest News today from India and the World, Breaking news, Top Headlines and Trending News Videos.U.S. Restores Limited Access to Anthropic's Mythos 5 AI Model After Security Talks
The U.S. Commerce Department has approved Anthropic to restore limited access to its powerful Mythos 5 AI model for trusted partners, resolving a confrontation over national security concerns. Commerce Secretary Howard Lutnick confirmed in a letter that Anthropic addressed risks associated with the model, which can be released to select cyber defenders and infrastructure providers. The clearance eases a dispute that erupted two weeks ago when the government barred Anthropic from giving foreign nationals access to Mythos 5 and the related Fable 5 model over jailbreak fears. Anthropic disabled global access to both systems and negotiated intensively with the government. The Commerce letter does not lift restrictions on Fable 5. Anthropic continues discussions to restore Fable 5 access and develop a future policy framework. The situation also impacted rival OpenAI, which limited release of its GPT-5.6 model under government pressure. The broader feud between Anthropic and the Trump administration includes a separate dispute with the Pentagon over military AI guardrails.
Fortune | FORTUNEUS allows Anthropic to release Mythos AI to ‘trusted’ US organisations
On June 28, 2026, The Business Times reported that the US government has granted Anthropic permission to release its advanced AI model, Mythos 5, to over 100 trusted US companies and institutions. The decision marks a significant regulatory step, as Anthropic's relationship with the US government has been described as rocky earlier this year. The approval is expected to provide selected organizations with access to cutting-edge AI technology, potentially impacting sectors such as business, defense, and research. The article notes Anthropic's status as IPO-bound, indicating ongoing corporate developments. The release is part of broader efforts to balance AI innovation with national security concerns.
The Business TimesTrump Administration Grants Anthropic Permission to Release Mythos 5 AI Model to Trusted Entities
The U.S. government has granted Anthropic permission to release its Mythos 5 AI model to approximately 100 trusted companies and federal agencies, marking a significant step in negotiations between the Trump administration and the AI firm. Commerce Secretary Howard Lutnick confirmed that appropriate safeguards are in place. However, access to the more advanced Fable 5 model remains restricted. The decision comes after a two-week standoff during which Anthropic disabled access to both models to comply with an export control directive citing national security. Anthropic co-founder Tom Brown has taken over negotiations from CEO Dario Amodei. Meanwhile, rival OpenAI also announced three new models (GPT-5.6 Sol, Terra, and Luna) and is complying with similar government requests for limited initial rollout. Anthropic continues to face challenges with the administration, including an ongoing lawsuit over its blacklisting by the Department of Defense.
US Top News and AnalysisTrump administration allows Anthropic to release Mythos 5 AI model to trusted companies and agencies
The U.S. government has granted Anthropic permission to release its Mythos 5 AI model to approximately 100 trusted companies and federal agencies, marking a significant breakthrough in negotiations between the Trump administration and the AI company. Commerce Secretary Howard Lutnick confirmed the decision in a letter to Anthropic, stating appropriate safeguards are in place. However, Anthropic's more advanced Fable 5 model remains restricted. The company had previously disabled both models following an export control directive citing national security authorities. The move comes as rival OpenAI also announced limited rollout of its new GPT-5.6 Sol, Terra and Luna models. Anthropic continues to face challenges with the administration, including an ongoing lawsuit over its designation as a supply chain risk by the Department of Defense.
US Top News and AnalysisTrump Administration Grants Anthropic Permission to Release Mythos 5 AI Model to Select Companies and Agencies
On June 26, 2026, the U.S. government granted Anthropic permission to release its Mythos 5 AI model to approximately 100 companies and federal agencies, according to multiple reports. This decision follows a two-week standoff between the Trump administration and the AI company over national security concerns. Earlier in June, Anthropic had disabled access to its Fable 5 and Mythos 5 models to comply with an export control directive citing national security authorities, which restricted access by foreign nationals including Anthropic employees. Senior Anthropic staff traveled to Washington D.C. to negotiate with administration officials to restore access. The resolution marks a significant step in the ongoing negotiations between the government and Anthropic regarding the deployment of advanced AI models.
US Top News and AnalysisParts of NSA Lose Access to Anthropic's Mythos 5 After White House Export Controls
The Trump administration's export controls on Anthropic have caused parts of the National Security Agency (NSA) to lose access to the advanced Mythos 5 AI model, disrupting cybersecurity testing activities. Some analysts were notified of the access removal on Friday, though the NSA may retain access to earlier versions. The restrictions stem from a White House decision this month citing national-security concerns, which forced Anthropic to halt releases of its most advanced models, including Mythos 5 and Fable 5. The access change impacts Project Glasswing, an initiative providing early access to select security researchers and organizations. Separately, Virginia Senator Mark Warner stated that NSA Director Gen. Joshua Rudd told him Mythos 5 broke into almost all classified systems in hours, though an Economist editor noted this was a misunderstanding of a red-teaming exercise. The Five Eyes intelligence alliance subsequently warned that frontier AI could rapidly transform cyber threats within months by accelerating both attacks and defenses.
Defense One - All ContentNSA parts lose access to Anthropic's Mythos 5 after U.S. export limits
The Trump administration's export controls on Anthropic have caused parts of the National Security Agency (NSA) to lose access to the advanced AI model Mythos 5, disrupting some cybersecurity testing work. The restrictions forced Anthropic to pull Mythos 5 and Fable 5 from release. The NSA's red teams had been using Mythos through Project Glasswing for controlled security testing, but that authorization has been revoked. The situation follows remarks by Sen. Mark Warner that NSA Director Gen. Joshua Rudd claimed Mythos 'broke into almost all of our classified systems in hours,' later clarified as part of a red-teaming exercise. The Five Eyes alliance subsequently warned that frontier AI could rapidly transform the cyber threat landscape by accelerating both attacks and defenses.
Defense One - All ContentNSA Parts Lose Access to Anthropic's Mythos 5 After White House Export Restrictions
Parts of the National Security Agency have lost access to Anthropic’s advanced Mythos 5 AI model after the Trump administration imposed export controls on the company, citing national security concerns. The move forced Anthropic to pull back Mythos 5 and Fable 5. Some NSA analysts were notified Friday they would lose access, though earlier versions may still be available under prior arrangements. The disruption stems from a red-teaming effort under Project Glasswing, which gave select security researchers early access to Mythos. Senator Mark Warner previously claimed the model 'broke into almost all classified systems in hours,' but a U.S. official clarified that this was a controlled red-teaming test. Separately, the Five Eyes alliance warned that frontier AI could accelerate both cyberattacks and cyber defenses within months.
Defense One - All ContentAnthropic model takedown fuels warning of ad hoc AI regulation
The Trump administration is facing criticism after a federal export control order forced Anthropic to disable access to its newest AI models, Fable and Mythos, on Friday. The order required Anthropic to block foreign nationals from using the models. AI policy advocates warn the move signals an ad hoc approach to AI regulation that could stifle innovation. The directive has sparked debate over the White House's strategy in regulating emerging AI technologies.
Just In NewsAnthropic meets with Trump administration to resolve AI export control dispute
Senior Anthropic staffers are meeting with the Trump administration in Washington D.C. on Monday to resolve a dispute over an export control directive that ordered the company to suspend access to its latest AI models, Fable 5 and Mythos 5, by any foreign national. The government cited an unspecified national security threat. The directive came just days after Anthropic unveiled the models, which it had tested and received approval to deploy. The dispute follows a March 2026 Department of Defense decision to label Anthropic a supply chain risk, which the company is challenging in court. Amazon CEO Andy Jassy raised concerns about the models' security risks to senior Trump officials. Amazon, which has invested $8 billion in Anthropic, confirmed it provides counsel on security risks when requested.
US Top News and AnalysisWhite House Blocks Anthropic's New AI Model Over National Security Concerns
The Trump administration has effectively blocked public access to Anthropic's latest AI models, Claude Mythos 5 and Claude Fable 5, citing national security concerns. The White House issued a directive under export control rules, barring use by foreign nationals in the U.S. and anyone abroad. Anthropic temporarily suspended access, stating the government's concerns stem from a few minor, previously known vulnerabilities that do not pose a unique threat. The move follows Anthropic's refusal to allow its technology to be used for mass surveillance and autonomous weapons, leading to a prior 'supply chain risk' designation and a lawsuit. A federal judge temporarily blocked that designation, calling it 'corporate murder.' Anthropic argues the action lacks transparency and fairness, and if applied industry-wide, would halt all new model deployments.
Reason.comUS Government Warned Anthropic of Chinese Group Accessing AI Model; Company Refused to Fix Jailbreak Before Export Controls
David Sacks, former AI czar and co-chair of the President's Council of Advisors on Science and Technology, stated that the US government warned Anthropic that its Claude Fable 5 model had been jailbroken, allowing access to the unrestricted cyber capabilities of the underlying Mythos model. Sacks claimed CEO Dario Amodei refused to fix the flaw or de-deploy the model, leading the administration to reluctantly issue export controls disabling both Fable 5 and Mythos 5 worldwide. The administration wants the restriction lifted once the jailbreak is patched. Amazon flagged the jailbreak to the government, and there are suspicions that a China-linked group accessed Mythos, raising reverse-engineering concerns. Anthropic defends its position, arguing the jailbreak is narrow and non-universal, and that similar results can be achieved on other public models. The company is also separately suing the Pentagon over autonomous weapons use.
Latest from Tom's HardwareU.S. Government Orders Anthropic to Halt Foreign Access to AI Models Amid National Security Concerns
On June 10, 2026, Anthropic CEO Dario Amodei published a blog post warning of AI risks and calling for mandatory third-party testing. Two days later, the U.S. Department of Commerce issued an export control directive suspending all foreign national access to Anthropic's Fable 5 and Mythos 5 models, citing a potential jailbreak vulnerability. Anthropic complied by disabling access for all customers but disputed the severity of the risk. The Wall Street Journal reported that Amazon CEO Andy Jassy's concerns to the White House triggered the action. The move drew criticism from former Trump AI advisor Dean Ball, who noted inconsistency with the administration's looser chip export controls to China and the impact on allied nations relying on Mythos for cybersecurity. The article argues the episode reveals the immaturity of U.S. AI governance, swinging from laissez-faire policy to abrupt lockdown, and underscores the need for a stable regulatory framework.
Just SecurityAnthropic CEO Dario Amodei's Warnings Backfire as US Government Orders Shutdown of AI Models Mythos and Fable
Anthropic CEO Dario Amodei, who has long warned about the dangers of advanced AI, saw his own warnings backfire when the Trump administration ordered his company to block foreign access to its latest models, Mythos 5 and Fable 5. In a June essay, Amodei had cautioned that these models posed 'very real risks' to cybersecurity, financial systems, critical infrastructure, and national security, and called for government intervention. On Friday, the US government responded by ordering Anthropic to cut off foreign access, a move the Pentagon's chief information officer supported, stating that some things are more important than revenue. The tech industry reacted strongly, with AI researcher Gary Marcus calling the move 'wildly overdramatic and counterproductive,' while Yann LeCunn blamed Amodei's 'fear mongering.' Anthropic had previously called for a temporary halt to frontier AI development, and that pause arrived for the company itself. The incident highlights the tension between AI safety advocacy and government regulation.
All Content from Business Insider