OpenAI Pledges $1 Billion for Cybersecurity Support After PHP Flaw Discovery

OpenAI has committed $1 billion to cybersecurity support after saying a frontier AI model identified a serious vulnerability in PHP before attackers did, putting the discovery alongside a major funding pledge.

OpenAI assessed the vulnerability as one of the most consequential security incidents in recent years. It tied that assessment to the flaw’s potential severity and the broad use of PHP applications. OpenAI also said its model found the weakness ahead of attackers, making the reported timing central to the announcement.

The funding will go through OpenAI’s Daybreak project. The initiative will provide grants to front-line cybersecurity defenders and support their use of Daybreak-related services. The commitment pairs assistance for defense workers with access to services connected to the project, while the reported discovery highlights a potential use for frontier AI in finding software weaknesses.

OpenAI linked the scale of its response to both the potential impact of the PHP flaw and the wide use of applications built with the language. The announcement presents early detection and Daybreak support as connected parts of its cybersecurity effort.