SAN FRANCISCO: Anthropic will soon begin restoring access globally to its most powerful AI models, Fable 5 and Mythos 5, after the US government lifted a restriction on where they could be released, the company said Tuesday. Over the past couple of weeks, the Trump administration has invoked national security concerns to limit the ability of major US tech companies to release advanced models, including those from Anthropic, which some researchers feared could be exploited to bypass cybersecurity measures. &quot;We&#39;ve received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5,&quot; Anthropic posted on X. &quot;We&#39;ll begin restoring access tomorrow.&quot; Just four days ago, the company said it had received authorization from the government to allow a small group of American cybersecurity firms to access Mythos 5. Secretary of Commerce Howard Lutnick said in a June 26 letter to the company that &quot;Anthropic has worked with the US government to address risks associated with the Covered Models,&quot; Politico reported. The government abruptly forced Anthropic to cut off access to its two cutting-edge artificial intelligence models on June 12 after discovering vulnerabilities in the safeguards put in place to prevent misuse of the tool. On Tuesday, Lutnick told Anthropic in a letter that the Trump administration had &quot;withdrawn&quot; its previous restrictions on the release of the company&#39;s models, Politico reported. The letter indicated that the Trump administration was satisfied, at least for now, that Anthropic had &quot;taken steps in close coordination with the U.S. government to address the risks associated with Claude Mythos 5 and Claude Fable 5.&quot; Like Anthropic, rival AI lab OpenAI has also complied with Washington&#39;s requests to restrict its own release of a new, powerful model called GPT-5.6 to a limited set of approved partners. &quot;This isn&#39;t quite the process that we think is optimal,&quot; OpenAI CEO Sam Altman said Friday in a post on X, alongside an explanation of the GPT-5.6 launch. Anthropic did not immediately respond to a request for comment Tuesday. However, in a blog post published Tuesday evening, Anthropic called for the development of a standardized &quot;framework&quot; to both assess critical vulnerabilities in advanced models and respond to them. The San Francisco-based AI lab will work with Amazon, Microsoft, Google and others on the effort. &quot;This problem will become more acute in the coming months, as more models with powerful cybersecurity (and other) capabilities are trained, assessed, and released,&quot; the blog post said. New frontiers&nbsp;The Trump administration issued an executive order on June 2 calling for the federal government to take multiple steps over the subsequent two months to take actions on AI and cybersecurity -- including creating a voluntary &quot;framework&quot; for private companies, such as Anthropic and OpenAI, to test and release their powerful &quot;frontier&quot; AI models in collaboration with the government. Susie Wiles, the president&#39;s chief of staff, posted Tuesday on X that the Trump administration was grateful for the cooperation from tech companies, though she didn&#39;t name any. &quot;My gratitude to companies across industries who continue to work closely with the White House to implement the President&#39;s&quot; executive order on AI and cybersecurity, Wiles said. &quot;This includes excellent work around advanced model access and guardrail testing and security.&quot; Earlier in the day, CIA Director John Ratcliffe compared the capabilities of the most advanced artificial intelligence models to nuclear weapons, in a tacit defense of the Trump administration&#39;s recent hard line on controlling the release of the most powerful AI technology. &quot;In conversations with many of the president&#39;s other national security and economic security advisors, we&#39;re talking about the impact of these frontier AI models,&quot; Ratcliffe said during a speech at the AWS summit in Washington. &quot;It would be...not misplaced to refer to their capabilities as akin to digital nuclear weapons,&quot; Ratcliffe said.

US President Donald Trump signed two orders on quantum computing accompanied by  US National Cyber Director Sean Cairncross (L), White House Office of Science and Technology Policy Director Michael Kratsios (3rd-L),  Commerce Secretary Howard Lutnick (3rd

Artificial Intelligence

"We've received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5," Anthropic posted on X. "We'll begin restoring access tomorrow."

US govt lifts restrictions on powerful AI models, Anthropic says

Faraz Saeed

LAHORE: Pakistan&rsquo;s technology and telecom sector unveiled the country&rsquo;s first AI-ready sovereign cloud, a move industry leaders said would fundamentally reshape national access to advanced artificial intelligence and end dependence on foreign cloud providers.&nbsp;The platform, jointly launched by major Pakistani tech and telecom firms, marked &ldquo;the moment Pakistan moved from consuming AI to producing it at scale,&rdquo; according to executives involved in the rollout.&nbsp;For years, enterprises seeking to develop high-performance AI models were forced to rely on overseas cloud services, constrained by global GPU shortages and the high cost of importing hardware. Local capacity was further limited by international export restrictions on advanced chips. Data Vault Pakistan, the country&rsquo;s AI-optimized and enterprise-secured data center, said it had overcome those barriers. The company secured special approval from NVIDIA for the import of its GPUs despite US restrictions targeting certain markets, enabling the deployment of thousands of high-performance processors inside Pakistan. Data Vault CEO Mehwish Salman Ali told Pakistan TV Digital that the company had expanded to &ldquo;over 3,000 GPUs,&rdquo; making it &ldquo;the only fully AI-enabled data center in the country.&rdquo; She called the milestone &ldquo;one of the hardest to achieve&rdquo; but said it now allowed banks, telecom operators, hospitals, and startups to train and deploy frontier-scale AI models domestically, with lower latency and full data sovereignty. &ldquo;The era of exporting Pakistan&rsquo;s most valuable data to foreign servers was over,&rdquo; Ali said. Ali added that control of GPU compute was as strategically important as national physical infrastructure. &ldquo;Whoever controls high-performance GPUs controls the speed of AI innovation,&rdquo; she said. &ldquo;Every rupee spent on foreign compute is a rupee that never comes back. Data Vault keeps that value and that power right here at home.&rdquo; She stressed that a sovereign AI cloud was &ldquo;not a buzzword, it&rsquo;s a red line,&rdquo; saying Pakistani data had to remain under domestic jurisdiction, something she argued global cloud providers &ldquo;cannot guarantee.&rdquo; The development coincided with a broader national technology push. In May, the Pakistani government allocated 2,000 megawatts of electricity in the first phase of a program to support Bitcoin mining and AI data centers. Officials said Pakistan&rsquo;s relatively low energy costs and available land gave it an advantage over regional competitors such as India and Singapore, where rising power prices and land scarcity constrained expansion. Demand for AI data centers surged globally, with consumption exceeding 100 gigawatts, while current supply was estimated around 15 GW. Pakistani industry leaders said that gap created an opening for countries with surplus energy and emerging regulatory frameworks.&nbsp;The new Telenor&ndash;Data Vault platform enabled organizations to run GPU-accelerated AI training, inference, and data processing. Supported uses included machine-learning pipelines, video analytics, generative AI systems, industrial automation, and the creation of large language models for Urdu and regional languages. Operators said sensitive datasets, including financial transactions, medical imagery, telecom records, and government information, would remain inside Pakistan, meeting compliance requirements set by the Pakistan Telecommunication Authority (PTA), the State Bank of Pakistan (SBP), health-sector regulations, and emerging national AI governance standards. The system was designed to support national priorities across multiple sectors. Financial institutions could strengthen fraud detection and anti-money laundering systems; hospitals could deploy AI diagnostics; manufacturers and logistics companies could use automation and predictive analytics; and government agencies could run secure, locally hosted AI models for public services and digital governance. The initiative also came as the telecommunications landscape underwent restructuring. Telenor Pakistan, with roughly 42.4 million mobile subscribers, was preparing to merge with PTCL, whose Ufone unit held 27.2 million subscribers. The combined entity was expected to form Pakistan&rsquo;s second-largest mobile operator, creating a broader network backbone for nationwide AI-driven services. Industry leaders said the convergence of domestic compute capacity, energy availability, strategic land allocation, and regulatory support positioned Pakistan to produce AI at scale rather than rely on international providers. The Telenor&ndash;Data Vault sovereign cloud, they argued, was both a technological breakthrough and a statement of national capability, signaling Pakistan&rsquo;s intent to build, control, and deploy advanced AI on its own terms.

The AI cloud ensures Pakistani data remains under local jurisdiction, enabling secure AI deployment for government, finance, and healthcare while safeguarding sensitive national information

Articles tagged with "AI models"

US govt lifts restrictions on powerful AI models, Anthropic says

Pakistan launches its first AI cloud service offering powerful processing capabilities

US govt lifts restrictions on powerful AI models, Anthropic says