After mocking Anthropic's Mythos rollout, OpenAI limits access to Cyber
What's the story
In a move similar to Anthropic's controversial release strategy for its cybersecurity tool Mythos, OpenAI has announced restricted access to its own cybersecurity platform, Cyber. The announcement was made by Sam Altman, who said that the company will start rolling out GPT-5.5 Cyber "to critical cyber defenders" in the coming days.
Tool features
Cyber toolkit can conduct penetration testing and identify vulnerabilities
The Cyber tool is capable of penetration testing, vulnerability identification and exploitation, and malware reverse engineering. As per the application on OpenAI's website, users have to share their credentials and intended use of the tool to gain access. The main aim behind this toolkit is to help companies identify security loopholes and test their defenses against potential cyber threats.
Misuse fears
Concerns over potential misuse of cyber toolkit
The introduction of the Cyber toolkit has raised concerns about its possible misuse by malicious actors. The tool's capabilities could be exploited if it falls into the wrong hands. This fear is similar to what was expressed when Anthropic restricted access to Mythos, with some critics even calling it fear-based marketing.
Wider availability
OpenAI addresses criticism, promises wider access
In response to the criticism, OpenAI has said that it is working on making Cyber more widely available. The company is consulting with the US government and looking for more users with legitimate cybersecurity credentials. This move comes after an unauthorized group reportedly gained access to Mythos despite Anthropic's restricted release strategy.