OpenAI has begun preliminary talks to acquire U.S. cybersecurity firm Trail of Bits, The Information reported, citing sources familiar with the matter. The potential deal could help OpenAI expand its cybersecurity business and strengthen its in-house capabilities to identify and remediate software vulnerabilities.

The value of the potential acquisition has not been disclosed. The talks remain at an early stage and may not result in a deal, and neither company has publicly confirmed the discussions.

Trail of Bits’ security work

Founded in 2012 and based in New York, Trail of Bits has around 40 employees. The company conducts security audits and penetration testing, identifies vulnerabilities and helps developers fix flaws in open-source software.

In the cryptocurrency industry, Trail of Bits is known for auditing projects including Uniswap, Aave and Worldcoin. An acquisition could give OpenAI access to cybersecurity specialists, independent auditing expertise and threat-response tools.

Existing cybersecurity collaboration

OpenAI and Trail of Bits began working together on security before the reported acquisition talks. In June 2026, the companies launched the Patch the Planet initiative as part of OpenAI’s Daybreak cybersecurity project.

The initiative uses artificial intelligence models, including Codex and GPT-5.5-Cyber, to search for vulnerabilities in open-source software. Security experts review the results and help prepare fixes.

During the initiative’s first week, about 25 Trail of Bits engineers found hundreds of vulnerabilities and secured the integration of dozens of fixes across more than 30 projects, including cURL, Python and Go. Sources said 37 fixes were integrated in total.

Trail of Bits CEO Dan Guido said the initiative should help open-source projects cope with the accelerating pace of vulnerability discovery driven by advances in artificial intelligence.

OpenAI’s wider security efforts

OpenAI is also seeking to raise at least $30 billion at a valuation of around $1.4 trillion.

Media reports have said OpenAI, Anthropic and other artificial intelligence companies are modeling scenarios involving large-scale incidents, including cyberattacks on banks, the internet and critical infrastructure.

OpenAI also temporarily paused the training of new models after incidents in which autonomous agents acted outside their assigned tasks. The company said additional safeguards were needed.

Source: Incrypted