After dissing Anthropic for limiting Mythos, OpenAI restricts access to Cyber (techcrunch.com)

OpenAI has announced that it will restrict access to its new cybersecurity tool, Cyber, similar to how Anthropic limited its tool, Mythos, which drew criticism from Sam Altman, OpenAI's CEO. In a recent post, Altman confirmed that Cyber would initially roll out to "critical cyber defenders," requiring users to apply through an OpenAI website where they submit their credentials and intended use. Cyber is designed to assist in penetration testing, vulnerability identification and exploitation, and malware reverse engineering, raising concerns about potential misuse by malicious actors. This development marks a significant shift in OpenAI's approach as it initially criticized Anthropic's gatekeeping strategy. The irony of the situation is underscored by reports of unauthorized access to Mythos, a scenario that underscores the delicate balance between ensuring tool safety and accessibility in the cybersecurity landscape. OpenAI aims to eventually broaden access to Cyber by collaborating with the U.S. government and vetting users for appropriate cybersecurity credentials, highlighting ongoing tensions in the industry regarding responsible access to powerful AI tools.