OpenAI CEO Sam Altman mocked Anthropic for restricting access to its cybersecurity tool Mythos. Now his company plans to limit its rival product, Cyber, in a similar way.
Altman made the announcement in a Thursday post on X. He stated OpenAI will start providing GPT-5.5 Cyber to critical cyber defenders over the next few days. Users must apply through a form on the OpenAI website. There they share details about their qualifications and how they plan to use the tool.
Background on the Tools
Cyber handles tasks like penetration testing, spotting vulnerabilities, exploiting them, and reverse engineering malware. The tool serves as a resource for organizations to detect weaknesses in their systems and check their protections. Concerns exist that malicious actors could abuse it for harmful purposes.
OpenAI launched in 2015 as a nonprofit research lab focused on safe artificial general intelligence. It shifted to a capped-profit model in 2019. Sam Altman became CEO in 2019 after serving as president. He faced a brief ouster in late 2023 before returning. The company gained fame with ChatGPT in 2022 and has released successive GPT models since.
Anthropic, founded in 2021 by former OpenAI executives including CEO Dario Amodei, emphasizes AI safety. It develops the Claude large language models. Mythos represents Anthropic's entry into AI-assisted cybersecurity.
Altman's Earlier Jab at Anthropic
Anthropic limited Mythos to approved users only. Altman labeled this approach as fear-based marketing. Other observers agreed, calling Anthropic's warnings excessive. In a twist, reports indicate an unauthorized group obtained Mythos access despite the controls.
OpenAI's decision mirrors Anthropic's strategy. This comes soon after Altman's public criticism. Both tools target cybersecurity professionals but carry risks if misused.
Path to Broader Access
Stay ahead of the AI curve
The most important updates, news, and content — delivered weekly.
No spam. Unsubscribe anytime.
OpenAI aims to expand Cyber's availability. It consults with the U.S. government on this. The company also verifies more users with valid cybersecurity backgrounds.
Cyber builds on OpenAI's GPT-5.5 foundation model. This version advances capabilities in specialized fields like security. The limited rollout prioritizes those defending critical infrastructure.
The cybersecurity field sees rising AI adoption. Tools like Cyber and Mythos promise faster threat detection. Yet dual-use potential demands caution. Governments and firms balance innovation with safeguards.
OpenAI's application process screens for legitimate needs. Applicants detail their roles in cyber defense. Approved users gain early access to test and refine the tool.
Anthropic's Mythos faced similar scrutiny. Its restrictions sparked debate on AI deployment in sensitive areas. OpenAI now navigates the same issues.
Industry Context
AI firms increasingly develop domain-specific models. Cybersecurity stands out due to high stakes. OpenAI competes with Anthropic, xAI, and others in this space.
Sam Altman's X post signals confidence in Cyber's value. He positions it as essential for defenders. The rollout timing follows Anthropic's move closely.
Both companies operate from San Francisco. They shape AI's role in national security. U.S. government input underscores the tools' importance.
Related on Neura Market