White House chief of staff Susie Wiles is scheduled to meet with Anthropic CEO Dario Amodei to discuss the San Francisco-based AI company's newly released Mythos model, a system so capable that it has drawn urgent attention from governments, regulators and financial institutions around the world. The meeting, first reported by Axios, comes as the administration seeks to evaluate advanced AI models for potential federal use, with a White House official emphasising that any technology considered for government deployment would first undergo a thorough technical review.
Anthropic announced the Mythos model on April 7, describing it as "strikingly capable" and warning that it surpasses human cybersecurity experts in identifying and exploiting software vulnerabilities. Because of those risks, the company has restricted access to a select group of customers. Alongside the release, Anthropic launched Project Glasswing, an initiative bringing together technology giants including Amazon, Apple, Google and Microsoft, as well as financial firms such as JPMorgan Chase, to help secure critical software infrastructure against the model's potential for harm. Anthropic co-founder and policy chief Jack Clark acknowledged at this week's Semafor World Economy conference that Mythos is not a unique development for long: similar systems from other companies are expected within months, and open-weight models from China with comparable capabilities could follow within a year to eighteen months.
The meeting takes place against a backdrop of significant tension between the Trump administration and Anthropic. President Trump moved earlier this year to bar all federal agencies from using Anthropic's Claude chatbot following a contract dispute with the Pentagon, while Defence Secretary Pete Hegseth sought to designate the company a supply chain risk — an unprecedented action against a domestic firm that Anthropic has contested in two federal courts. A federal judge blocked enforcement of Trump's directive in March. Despite that friction, even prominent Anthropic critic David Sacks, who served as the White House's AI and crypto adviser, urged people to take the Mythos threat seriously, saying its advanced coding abilities make it genuinely capable of chaining together software vulnerabilities into working exploits.
Concerns are spreading well beyond Washington. The United Kingdom's AI Security Institute evaluated Mythos and found it a meaningful step forward over already rapidly improving predecessors, warning that it can exploit systems with weak security and that more models with similar capabilities are likely to emerge. Barclays CEO C. S. Venkatakrishnan, speaking on the sidelines of the IMF's spring meeting in Washington, described Mythos as "a serious issue" for global banking, noting that legacy systems at large financial institutions could be particularly vulnerable. He cautioned that Mythos is unlikely to be the last such model: "There will be a Mythos 2 and a Mythos 3, and they'll come with probably distressing frequency." The European Commission confirmed separately that Anthropic has been in talks with EU officials about its AI models, including advanced systems not yet released in Europe.
The broader debate over how to govern frontier AI is being thrown into sharp relief by Mythos. The model encapsulates the central dilemma facing policymakers: systems capable enough to help defend critical infrastructure are, by definition, also capable enough to attack it. How governments and the private sector coordinate their response — and how quickly — may determine whether the technology's defensive potential outpaces its risks.