How did AI discover the internet's massive cyber-security risk?

Governments and cyber-security teams around the world are rushing to deal with a serious weakness in internet technology. AI found problems that hackers could use to attack computer systems.

The developers of PHP, a technology used by more than 70% of websites, including Wikipedia and online shopping sites, released updates to fix security problems. These weaknesses had been hidden in the software for years.

Cyber-security company Wiz called the problem the "most severe" ever found. OpenAI, whose technology was used to discover the weaknesses, called the discovery "one of the most impactful security issues in recent memory".

But how did AI find such a serious security risk affecting the internet, and what does it mean for you? The BBC's technology correspondent Joe Tidy reports.