AI Discovers Two Critical Vulnerabilities in PHP with CVSS Score of 9.2

The Artificial intelligence It is here to stay, and to help us in many of our daily tasks. This includes the security flaw detection in systems widely used by users, such as the most recently discovered in PHP. There have been two vulnerabilities found with a score CVSS of 9.2 that affect a multitude of systems, making it essential to apply patches to the affected versions according to the PHP Foundation itself.

AI Discovers Two Critical Vulnerabilities in PHP with CVSS Score of 9.2

The initiative Patch the Planet has been responsible for discovering these critical flaws in PHP, operated by software maintainers, researchers and cybersecurity companies. One of the bugs has been discovered by GPT-5.6 Sunwhile Wiz also managed to detect the two vulnerabilities independently. To launch the corresponding patches, we have had the collaboration of PHP maintainers, with the technical support of OpenAI, Wiz, Cloudflare and other important partners.

To act against this type of vulnerabilities, 5 important steps.

1. Validate and correct the failure, buying its impact and testing the solution against possible attempts to evade it.
2. Notify those interested and affected in advance. This way they can prepare updates and other protective measures before the vulnerability becomes known.
3. Coordination with different public organizations to protect the most critical infrastructures and maintain essential services.
4. Prepare updates and mitigations for all affected infrastructures.
5. Publish recommendations and deploy patches, with help identifying vulnerable systems and verifying that updates have been applied correctly.

OpenAI will allocate $1 billion to improve security with AI

Besides, OpenAI has stated that it will allocate $1 billion to subsidize access to Daybreak through a program aimed at organizations with limited security resources. It will also offer financial aid, access to AI models and technical assistance to help maintainers of Open Source projects.

But this can also benefit users with bad intentions, so they are working on limit malicious use of tools without security investigations being able to be carried out.

OpenAI proposes using AI to develop more secure software

OpenAI also wants you to use the Artificial Intelligence to improve software security from its developmentproposing habits such as safe default programming, migration to memory-safe languages, and reinforcing isolation, among others. Researchers and other users can share their suggestions on the different proposals put forward by OpenAI.