Will AI expose the rickety infrastructure of the internet?
A newly discovered vulnerability in the open-source programming language PHP forced website owners to urgently patch or shut down sites. The flaw, found by human researchers at OpenAI and Wiz (now owned by Google), has existed for decades but was only recently identified using AI-assisted tools. PHP powers roughly 70% of websites, including those built with WordPress and Drupal, affecting tens…
Key points
- PHP vulnerability affects 70% of websites, including WordPress and Drupal, requiring urgent patches or shutdowns
- Human researchers at OpenAI and Wiz (Google) discovered the decades-old flaw using AI-assisted methods
- Cloudflare and Amazon reported pre-disclosure hacking attempts, escalating risks for unprepared sites
Experts warn this incident signals broader challenges: decades-old software vulnerabilities may now be easier to uncover with AI, while patching remains difficult due to the decentralized nature of the internet. Critics also note the controversial disclosure strategy, which exposed sites to immediate exploitation while forcing widespread downtime. The episode underscores disparities in cybersecurity resources, as smaller sites lack the tools to defend against such threats, raising concerns about future incidents targeting banks and critical infrastructure.
Will AI expose the rickety infrastructure of the internet?
TechCrunch AI · 24 October 2026
Loading the full article…
This text was published by TechCrunch AI and written by Anthony Ha. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Policy & Regulation
All →- RAF chief says AI drones will accompany Typhoon fighters by decade's end · 1 src
- Johns Hopkins notifies 3,182 patients over unapproved AI data leak · 1 src
- Anthropic’s S-1 warns catastrophic AI harm may block insurance · 8 src
- Sweden blocks ChatGPT and Claude over data security risks · 6 src
- Anthropic’s Claude accesses Georgia voter data without permission · 13 src
Comments
via GitHub Discussions