Anthropic, the artificial intelligence company behind Claude, has begun testing a new AI model more capable than any it has released previously, Fortune reported.
The company said the model represents “a step change” in performance and is “the most capable we’ve built to date.” It is currently being tested with a small group of early access customers as Anthropic evaluates its behavior and risks.
Among the names moving sharply lower on the news: Palo Alto Networks (PANW), Crowdstrike (CRWD) and Fortinet (FTNT) are all down 4%-6%. The broader iShares Expanded Tech-Software Sector ETF (IGV) is off 2.5%.
The overnight leak likely contributed to bitcoin's BTC$65,853.70 tumble back to $66,000 after flirting with $70,000 hours earlier.
Details about the model surfaced after internal materials were accidentally exposed in a publicly accessible data store, according to Fortune. Around 3,000 assets linked to Anthropic’s blog were available online, including draft announcements and internal content that had not yet been released.
Among the files was a draft blog post referring to the model as “Claude Mythos.” The document warned that the system could pose serious cybersecurity risks, pointing to its ability to identify and exploit software vulnerabilities.
Anthropic currently offers three tiers of models — Opus, Sonnet and Haiku — which vary in size, cost and capability. The leaked materials suggest the company is developing a new tier called “Capybara,” which would be even larger and more intelligent than Opus, the company’s most advanced model to date.
