Anthropic has completed training a new model — the successor to Mythos 5
The artificial intelligence market has received an important signal: according to analyst Andrew Curran, the company Anthropic has completed training a new model that could become the next generation of the Mythos line. This system reportedly surpasses Mythos 5 in capabilities, but there has been no official confirmation from the company itself — neither a name, nor stated specifications, nor even confirmation of the model's existence.
What is known about the new model?
Curran notes that training is complete, but the system's fate remains uncertain. It could be released under the name Mythos 5.1 or Mythos 6, or it might be kept entirely for internal use. This announcement came nine days after US export restrictions forced Anthropic to suspend work on Mythos 5 and Fable 5.
As a reminder: on June 12, US Secretary of Commerce Howard Lutnick sent a directive to Anthropic CEO Dario Amodei, citing national security concerns. The restrictions affected all foreign nationals, including company employees born outside the US. As a result, Anthropic disabled both models worldwide.
Why didn't the suspension stop development?
Notably, the development block not only failed to slow progress but, according to Curran, may actually accelerate it — by freeing up resources. He cites open-source competitors such as Z.ai's GLM-5.2, which handles programming tasks just as well as much more expensive closed models.
Anthropic continues to seek the lifting of export restrictions. The company claims its researchers used Fable 5 to search for information with potentially dangerous applications, and this incident was even commented on by Donald Trump. However, the government's decision remains in effect, setting a significant precedent for the entire industry.
The market awaits
Whether Anthropic releases the new model publicly, restricts access through the Glasswing project, or keeps it for internal use — this will fundamentally change the balance of power in the AI market. To date, about 50 partners using early versions of Mythos have identified over 10,000 critical vulnerabilities in software. Clearly, the system's potential is enormous, and its future fate will serve as an indicator of how regulatory pressure affects innovation in the field of artificial intelligence.
My analysis: The situation surrounding Anthropic demonstrates a classic conflict between safety and progress. The blocking of Mythos 5 and Fable 5 did not stop development — it merely redirected resources. The new successor, even if it remains an internal tool, will give the company a critical advantage. However, what matters more for the market is whether Anthropic can get the restrictions lifted and regain access to its flagship models. This will determine who sets the rules in the next cycle of the AI race.