> ## Content Index
> Fetch the complete content index at: https://www.theleftshift.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# OpenAI Plans Limited Release of Advanced Cybersecurity AI Amid Rising Threat Concerns
- URL: https://www.theleftshift.com/openai-plans-limited-release-of-advanced-cybersecurity-ai-amid-rising-threat-concerns/
- Published: 2026-04-10T05:37:16.000Z
- Updated: 2026-04-10T05:37:16.000Z
- Description: The development comes as rival Anthropic recently announced a restricted rollout of its Mythos Preview model to select technology and cybersecurity firms.
- Author: The Left Shift Bureau
- Tags: Cybersecurity

OpenAI is finalising a new AI product with advanced cybersecurity capabilities, expected to be released to a small group of trusted partners, *Axios* [reported](https://www.axios.com/2026/04/09/openai-new-model-cyber-mythos-anthopic?ref=theleftshift.com).

The development comes as rival [Anthropic](https://www.theleftshift.com/anthropic-unveils-claude-managed-agents-to-cut-ai-development-time-from-months-to-weeks/) recently announced a restricted rollout of its [Mythos Preview model](https://www.theleftshift.com/anthropic-launches-project-glasswing-with-tech-giants-to-tackle-ai-driven-cybersecurity-risks/) to select technology and cybersecurity firms due to fears around its hacking capabilities. 

OpenAI has taken a similar approach through its “Trusted Access for Cyber” pilot, launched earlier this year following the release of its [GPT-5.3-Codex model](https://www.theleftshift.com/openai-launches-codex-security-ai-agent-to-hunt-and-fix-software-vulnerabilities/).

The invite-only programme provides organisations with access to more permissive and cyber-capable models that support defensive security work. [OpenAI](https://www.theleftshift.com/openai-acquires-tbpn-to-expand-ai-communication-and-industry-engagement/) has also committed $10 million in API credits to participating groups.

[Security experts warn](https://www.theleftshift.com/ai-is-compressing-cyberattacks-to-minutes-booz-allen-warns/) that AI has reached a point where it can identify vulnerabilities and potentially exploit them at scale. Concerns are mounting that such tools, in the wrong hands, could target critical infrastructure such as power grids, water systems, or financial networks.

Despite efforts to limit access, experts say the capabilities cannot be contained indefinitely. Industry leaders compare the staggered release of such models to responsible vulnerability disclosure practices in [cybersecurity](https://www.theleftshift.com/acronis-launches-24-7-mdr-service-to-help-msps-scale-cybersecurity-without-in-house-soc/), though they caution that similar capabilities may soon become widely accessible.