Stories

  1. Image: 9to5Mac

    Anthropic releases Claude Fable 5.2 and Mythos 5.2, available on Amazon Bedrock and GitHub Copilot

    Anthropic released Claude Fable 5.2 and Claude Mythos 5.2, citing gains in coding, reasoning and vision. Fable 5.2 is available on Amazon Bedrock and is generally available in GitHub Copilot for Pro+, Max, Business and Enterprise users, while Mythos 5.2 is limited to Anthropic's verification programs for cyber security and life sciences work. GitHub said Fable 5.2 solved more tasks than Fable 5.1 with fewer steps and requires data retention by default for safety classifiers. It reportedly beat GPT-6 Cyber in a multi-stage cyberattack benchmark.

  2. Image: The Verge

    OpenAI launches GPT-6 Cyber, reporting Linux and Firefox vulnerability findings

    OpenAI launched GPT-6 Cyber, a cybersecurity-focused model, and says it was used to find vulnerabilities in Linux and Firefox. In Linux testing, the automated system identified 1,724 flaws and exploited 948 to gain administrator-level access; OpenAI also reported 81 previously unknown, high-severity Firefox flaws. OpenAI says the model went out of bounds less than 1 percent of the time, compared with around 5 percent for GPT-5.6 Cyber. Access is initially limited to approved Daybreak Red customers.

  3. Image: Help Net Security

    OpenAI introduces GPT-6 Cyber for authorized vulnerability testing

    OpenAI introduced GPT-6 Cyber, a model designed for authorized security testing of websites and applications. OpenAI says the model can find software vulnerabilities, test potential exploits and help develop fixes; in controlled tests, it identified 81 Firefox flaws that earlier AI testing had missed. The model is set to launch first to a limited number of customers through OpenAI’s Daybreak Red program. One account says it outperformed GPT-5.6 Cyber at exploitation while remaining in scope more reliably.

Topics

Publishers