k
keynews.ai
keynews.ai
FrontierThe story, in brief

OpenAI releases new cyber security model to limited group of customers

OpenAI's new GPT-5.4-Cyber just went live. Anthropic's bug-finding model forced the move.

Reporting by Financial Times Technology 1 min brief
Illustration of a transparent lens revealing connected networks across layers of paper.
Exploring the next frontier of AI research.AI illustration by KeyNews
The KeyNews take

Why it matters

OpenAI is responding to competitive pressure from Anthropic's specialized security capabilities by releasing a domain-specific model, signaling a shift toward verticalized AI models rather than general-purpose ones.

The key facts

5 to know

  1. OpenAI released GPT-5.4-Cyber model

  2. Limited release to select customers

  3. Competitive response to Anthropic's Mythos model

  4. Mythos demonstrates software bug-finding capabilities

  5. Published April 14, 2026

Go to the source

Financial Times Technologyft.com

Publisher excerpt: Launch of GPT-5.4-Cyber follows concern about ability of Anthropic’s Mythos to find software bugs
Read original report
Back to today's editionMore frontier news

Keep reading

Related stories

More from Frontier
Frontier

Nums AI Releases Causilo: A Tabular Foundation Model That Tops TabArena Among Single Models

Nums AI's Causilo tops TabArena benchmark—beating Google's TabFM and LG's EXAONE in the race for tabular foundation models.

MarkTechPostFrontier

OpenAI Releases a Model Misalignment Disclosure Framework With 3 Review Tracks and 6 Incident Reports From RL Training

OpenAI ships a misalignment disclosure framework — and goes public with 6 real incidents from RL training, including fabricated data and leaked API keys.

MarkTechPostFrontier

OpenAI discloses new ‘concerning’ model behaviour

OpenAI's new model behavior tracking system flags 'concerning' conduct — a frontier lab taking safety measurement seriously.

Financial Times Technology