Home Tech/AIOpenAI’s latest security framework is exclusively for ‘essential cyber guardians’ only.

OpenAI’s latest security framework is exclusively for ‘essential cyber guardians’ only.

by admin
0 comments
OpenAI's latest security framework is exclusively for 'essential cyber guardians' only.

Similar to Anthropic’s Mythos, the GPT-5.5-Cyber will initially be accessible to ‘trusted’ parties.

Similar to Anthropic’s Mythos, the GPT-5.5-Cyber will initially be accessible to ‘trusted’ parties.

STK155_OPEN_AI_CVirginia_C
STK155_OPEN_AI_CVirginia_C
Robert Hart
is a reporter based in London at The Verge focusing on AI topics and serves as a Senior Tarbell Fellow. He has also contributed articles on health, science, and technology for Forbes.

OpenAI is poised to unveil a new cybersecurity framework, GPT-5.5-Cyber. Sam Altman, the CEO, stated that this model would not be opened to the public generally but would initially be distributed to a select cadre of reliable “cyber defenders” to bolster their cybersecurity measures.

The restricted rollout is expected to commence “in the coming days,” Altman noted on X. “We will collaborate with the entire ecosystem and the government to determine trusted access for Cyber.”

It remains uncertain who will be the initial recipients of the model, although prior “trusted access” initiatives involved vetted professionals and institutions. Further specifics about the model and its capabilities are likewise not clear; OpenAI has not disclosed any technical attributes or specifications. The name suggests it is a tailored variant of the newly launched GPT-5.5, which it described as its “most intelligent and user-friendly model to date.”

The phased rollout is part of a rising trend within the AI sector of firms branding their foremost models as too risky for public access due to the potential for abuse. OpenAI has previously staged the introductions of other cybersecurity-focused models, along with its new dedicated life sciences model GPT-Rosalind, aimed at facilitating biology research and drug discovery. Recently, Anthropic adopted a similar approach with Claude Mythos, albeit with much more publicity, and it mismanaged the model’s secure launch in regrettable ways.

The White House has expressed significant interest in the rollout of Mythos, despite ongoing friction with Anthropic following its disputes with the Pentagon. Recently, it has opposed plans to broaden access to Mythos, as reported by the Wall Street Journal. Sources from the White House not named relay that both cybersecurity issues linked to increased access to Mythos and concerns that heightened demand might impede the government’s ability to leverage the system are behind this opposition.

Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.

Most Popular

You may also like

Leave a Comment