Tech & Science In the Wake of Anthropic's Mythos, OpenAI Has a New Cybersecurity Model—and Strategy
OpenAI says its safeguards “sufficiently reduce cyber risk" for now, while GPT-5.4-Cyber is a new cybersecurity-focused model.
By
· April 14, 2026 · 1 min read
This article was originally published by
Wired
and is republished here under license.
OpenAI says its safeguards “sufficiently reduce cyber risk” for now, while GPT-5.4-Cyber is a new cybersecurity-focused model.
Like this:
Like Loading...
Related
Strong coverage and a long run time make this pool-cleaning robot a compelling alternative to pricier models.…
By
· April 22, 2026 · 1 min read
A WIRED review of permits for data center projects using natural gas and linked to OpenAI, Meta, Microsoft, and xAI…
By
· April 22, 2026 · 1 min read
On May 8, we’re going live to answer your questions about the Musk v. Altman trial that could determine the…
By
· April 22, 2026 · 1 min read
Leave a Reply