DistantNews
Support us
🇰🇷 South Korea /Technology

[Weekly AI] From performance to safety: Anthropic, Google and OpenAI roll out new models

From Dong-A Ilbo · () Korean

Translated from Korean and summarized by DistantNews. Read the original for the full story.

At a glance

In-depth From a news agency New plan
  • Anthropic released Claude Mythos 5.1 and Claude Fable 5.1 on Sept. 1, using the same base model with different safety protections.
  • Fable 5.1 is available to the public through an API, while Mythos 5.1 is restricted to invitation-only programs for cybersecurity and life-science experts.
  • The company emphasized safety alongside performance improvements.

Artificial intelligence is becoming inseparable from everyday life, and the major technology companies are continuing to release new models. This week, Anthropic put safety as well as performance at the center of its latest announcement.

Anthropic released Claude Mythos 5.1 and Claude Fable 5.1 on Sept. 1. The two models use the same underlying model, but Anthropic applied different levels of safety safeguards to them.

Fable 5.1 will be made available to the general public through an API. Mythos 5.1 will take a more restricted route, offered only through invitation-based programs for specialists.

Those programs include the Cyber Verification Program, or CVP, for cybersecurity experts, and the Life Sciences Verification Program, or LSVP, for life-science experts. Anthropic said its announcement focused not only on improving model performance, but also on safety.

About this summary

Originally published by Dong-A Ilbo in Korean. Translated, summarized, and contextualized automatically by DistantNews, with a note on how the source frames the story. Not individually reviewed before publishing. How this works.

More Stories

From Our Blog