After PHP-related outages hit government sites, cybersecurity experts say AI is uncovering software flaws faster than ever. Here’s what it means for defenders and website owners.
AI-Discovered Software Bugs Could Trigger More Outages, Cybersecurity Experts Warn
Artificial intelligence is now finding software flaws faster than the industry can comfortably fix them. According to seven cybersecurity professionals who spoke to Reuters, that is almost certain to cause more disruption like the website outages seen across government sites this week.
What Happened With PHP?
The latest scare involves PHP, an older programming language that runs a huge share of the web. Starting Wednesday, researchers disclosed weaknesses in PHP, and operators raced to patch their systems before attackers could exploit them. Government websites in Europe and elsewhere were among those affected.
Some site owners felt they couldn’t patch in time, so they simply took their sites offline until they could secure them. That is a drastic step, and it shows how seriously organizations took the threat.
A Busy Week for Security Teams
PHP wasn’t the only story. VulnCheck researcher Patrick Garrity noted that the same week also brought:
- Vulnerability disclosures affecting Node.js, a PHP alternative, and the streaming software Plex
- Breaches tied to the AI gateway LiteLLM and the development platform Railway
- A report that Indian certificate authority eMudhra had used an extremely weak cryptographic key, potentially putting users across the internet at risk
Garrity’s view is that if this many incidents land in a single week, more are coming, and anything exposed to the internet could face repeated emergency “fire drills.”
The Numbers Behind the Trend
VulnCheck, a Massachusetts-based threat intelligence firm, is due to publish research in the coming days. Its findings say AI helped roughly double the number of digital flaws discovered compared with a year earlier.
Cloudflare CEO Matthew Prince told Reuters he knows of several more severe vulnerabilities in other critical software that are currently being patched. He expects a long stretch of nonstop alarms, saying that for a couple of years it may feel like “the sky is falling” every week.
Why the Timeline Is Shrinking
Software today is built from enormous codebases, and bugs can hide almost anywhere. Until recently, finding them took real skill, and turning a bug into a working attack tool (an “exploit”) could take an expert weeks or months. That gave defenders time to prepare and roll out fixes gradually.
AI has changed that. Models can scan code around the clock, uncover one weakness after another, and produce working exploits in hours. Shane Fry, CTO of RunSafe Security, said recent results have ended any real debate about whether AI will reshape the industry. The big shift, in his view, is how fast bugs are found and how quickly attackers can weaponize them.
The same trend is behind the growing need for fast patching and AI-assisted defense. If you want to see how AI is being used on the builder side, read our guide to the Top 5 AI Tools That Actually Help Developers in 2026.
Will More Bugs Mean More Hacks?
All the experts agreed that defenders have less time than before. They disagreed on whether this will lead to a wave of successful attacks.
- The skeptics: Daniel Stenberg, lead developer of the file transfer tool curl, argues that AI tends to find obscure bugs that are hard for criminals to use. Ramy Rahman of ArmorCode made a similar point: more disclosed vulnerabilities don’t automatically mean more successful attacks.
- The worriers: Prince said he has seen anecdotal evidence of compromises linked to the PHP flaw.
The early PHP evidence is mixed. Hunt.io, Amazon and Cloudflare all reported aggressive attempts to exploit the bug. But Hunt.io’s head of research, Esteban Borges, said that as of Thursday afternoon US Eastern time, the criminals his team tracks appeared not to have built a working tool yet.
What Website Owners Can Do
This section is our own general advice, not from the Reuters report:
- Patch quickly. Apply security updates for your server software, programming language and plugins as soon as they are released.
- Ask your host. If your site runs on PHP, including many popular content management systems, check whether your hosting provider has applied the latest fixes.
- Back up regularly. A recent backup makes recovery much easier if something goes wrong.
- Reduce exposure. Remove unused plugins, tools and services, and consider a web application firewall.
- Watch for alerts. Follow security advisories from your platform and hosting provider.
The Bottom Line
AI is helping defenders and attackers alike find flaws that have been hidden for years. Even if not every disclosed bug becomes a successful hack, the time between discovery and attack is shrinking. Companies and website owners that can patch fast will be in a much better position than those that can’t.
For more on how AI is changing technology and markets, browse our AI News and Tech News sections. You may also like our look at Nvidia, Oracle, CoreWeave AI Stocks Sink After OpenAI Revenue Report, and our article on Can You Trust an AI Financial Advisor?, which covers AI security and risk in personal finance.
Source: Reuters, October 23, 2026. This article is for informational purposes only.

