Best Miami News connects businesses and publishers

collapse
Home / Daily News Analysis / The internet runs on PHP. AI just found a big problem. Should you be worried?

The internet runs on PHP. AI just found a big problem. Should you be worried?

Oct 24, 2026  Twila Rosenbaum 48 views
The internet runs on PHP. AI just found a big problem. Should you be worried?

Key facts at a glance

  • Two critical PHP vulnerabilities, CVE-2026-109618 and CVE-2026-109619, affect all PHP installations that serve web requests and carry a severity rating of 9.2 out of 10.
  • Both flaws permit remote code execution, which means an attacker could run commands on a vulnerable server.
  • Emergency patches are available in PHP 8.5.12, 8.4.27, 8.3.36, and 8.2.35. Older versions need fixes from a vendor that maintains them.
  • Researchers developed a single-script exploit called PHP2Shell, and AI models helped uncover the flaws and shorten the path to working attacks.
  • Millions of exploit attempts were reported within hours, and some probing may have begun four days before the patches were released.
  • Because PHP supports a large share of the web, an unpatched site could be used to spread malware, steal data, or delete data.

What happened

On October 20, 2026, the PHP Group issued an alert for coming security updates designed to patch two very high-risk vulnerabilities. The flaws were later identified as CVE-2026-109618 and CVE-2026-109619. Both allow an attacker to run remote code, both carry a severity rating of 9.2 out of 10, and together they affect all PHP installations that serve web requests, according to the maintainers. The advisory set off urgent work across the web as server administrators prepared for patched versions of PHP. When details of the two vulnerabilities were released the next day, October 21, sites began going offline temporarily as upgrades were completed. Government websites in the Netherlands, France, and Germany were among those affected. The existence of the flaws had been known before, but they were assumed to be too difficult to exploit. Security researchers then found that the vulnerabilities could be triggered through a single script. That exploit was named PHP2Shell. The script was developed with help from frontier AI models, according to the researchers, and it changed the practical risk from theoretical to immediate. By the evening of October 21, one major content-delivery network reported that it had already blocked millions of attempted exploits. Its chief executive said the numbers spiked as soon as the flaw disclosures went live. Cloud-security researchers also found signs that attackers had been probing the flaws as early as October 17, four days before emergency patches landed. Attackers were reportedly adjusting their techniques to slip past defenses. The lesson is blunt: a website that runs PHP and remains unpatched should not assume it is safe just because nothing suspicious has been noticed. Administrators were advised to check for signs of compromise dating back to October 17.

Why PHP still matters

PHP is often treated as background infrastructure, but it remains one of the most consequential languages on the web. It is a general-purpose scripting language used heavily in web development. When it runs on a server, it lets websites handle complex actions without embedding every instruction or piece of content directly inside a page. That server-side work produces dynamically generated output: the current date on a calendar, customized HTML, PDFs, and even JSON or XML files used for caching and data exchange. The language dates to the 1990s, when it emerged from early web tooling and grew into a practical way to build interactive sites. Over the decades it has been refined, extended, and embedded in countless hosting stacks. Many websites use a server-side language to serve content, and industry estimates suggest that almost 70 percent of those sites rely on PHP. Across the broader internet, one widely used content-management system built on PHP is estimated to power around 40 percent of all websites. That scale explains why a PHP security problem becomes an internet-wide event. It is not because every site is identical, or because every administrator ignores updates. It is because PHP sits under so much of the web, including small business sites, publishing platforms, government portals, community groups, and hobby projects. A flaw in one widely deployed interpreter can become a flaw in millions of installations at once. For consumers, the result can be invisible: if administrators patch quickly, ordinary browsing may continue without interruption. For website owners, the result is a race against attackers who can scan, probe, and attempt exploitation at machine speed.

Why the AI angle changed the risk

The most important part of this story is not only that the vulnerabilities were severe. It is how quickly they became practical to exploit. In the pre-AI era, a known vulnerability could take days, weeks, or even months before a working exploit appeared. The spread of attacks was slower, and discovery often took longer. That timeline has compressed. In this case, AI did not simply accelerate the threat after disclosure. It helped researchers uncover the flaws in the first place. Researchers used frontier AI models while investigating the bugs and developing working exploits. According to their account, AI helped turn vulnerabilities that had lurked in PHP for more than a decade into something a single researcher could exploit in a week. The same task might previously have required a world-class security team and months of work. That is the double-edged sword of AI in security. Defenders can use powerful models to find dangerous bugs before attackers do. But the same capability rapidly lowers the time and expertise needed to turn obscure flaws into working attacks. Attackers can iterate faster, adjust payloads, and probe for gaps in firewalls and filtering rules. The defensive response has been substantial. Services erected firewalls and other protections to block exploit attempts, and one of the largest content-delivery networks reported blocking millions of attempts. But those defenses are not impenetrable. Coverage does not extend to every corner of the internet, and one major network covers only about 21 percent of all websites. Attackers continue to adapt. A firewall can reduce risk, but it cannot replace a patch. The combination of a widely deployed language, a critical remote-code-execution flaw, and AI-assisted exploitation is what made this week so urgent.

Which PHP updates include the security patches?

The PHP Group directly issued patches within versions 8.5.12, 8.4.27, 8.3.36, and 8.2.35. A website that relies on an older version must receive the necessary security fixes through a vendor that helps maintain that older branch. Updating is not always as simple as clicking a button. Some sites depend on custom extensions, old plugins, or bespoke code that must be tested against a new PHP release. That is why emergency patching can temporarily take sites offline. The safer path is still to patch quickly, test critical functions, and keep a rollback plan. Administrators should also confirm that their hosting provider has not merely updated a control panel while leaving the actual PHP runtime on an unsafe version. Managed hosts may handle the work quietly in the background. Self-managed servers require direct attention. That includes home servers, single-board computers, network-attached storage devices, and any other system running PHP as part of a web stack. If a site is down, it may be down because the administrator is completing an upgrade. If a site is up, it may still be running an unpatched version behind a firewall. The only reliable answer is to check the version and verify the patch.

What website owners should do

Website owners should first determine which PHP version their site is using. They should then update to one of the patched releases or confirm that their host or vendor has done so. After updating, they should check firewall and web-application firewall rules, review logs for unusual requests, and look for signs of compromise that may date back to October 17. Backups should be current and stored separately from the production server. Administrators should also review accounts, file permissions, and scheduled tasks for unexpected changes. If a site runs a common content-management system, its admin dashboard may show the PHP version. If the site is managed by a hosting provider, the owner should ask for written confirmation that the runtime has been patched and that exploit attempts are being blocked. Those running their own servers should apply the update directly, restart the relevant services, and test the site after the change. They should not wait for visible signs of attack. Remote-code-execution flaws can be exploited quietly, and a compromised server may continue to serve normal pages while attackers use it for other purposes. The reported probing before the patch release is a reminder that attackers do not wait for public disclosure. A site that was exposed before the update should be checked for webshells, modified files, unexpected scheduled jobs, and outbound connections to unknown hosts. If compromise is suspected, the site should be isolated, cleaned, and rebuilt from trusted backups rather than merely patched.

What visitors should know

Most visitors do not need to change their browsing habits dramatically. Many websites are operating normally, and many hosts are applying patches automatically. Some sites may go offline temporarily while upgrades are completed. The general advice remains sound: keep browsers and antivirus software up to date, share only necessary personal information with websites, and be cautious about downloading files from sites that behave strangely. If important data is stored on a website, keep backups of that data in case the service is disrupted. A compromised website could be used to spread malware, steal data from the server, or delete data on the server. Visitors cannot patch a remote site, but they can reduce their own exposure by using current software and avoiding unnecessary risk. Website owners, however, carry the main responsibility. They should verify their PHP version, apply the official patches, and check for compromise rather than assuming that a firewall is enough. The incident shows how quickly a deep infrastructure flaw can become a public-facing security event. It also shows why basic maintenance matters. The internet depends on a long chain of servers, languages, and content-management systems. When one link in that chain is critical, the fix has to travel through every site that depends on it.

How to check a site's PHP version

A hosting dashboard may show the full details of the PHP version in its general management area. If it does not, a site running a common content-management system may reveal the version inside its admin dashboard. In the left-hand navigation bar, look for Tools, then choose Site Health. Two tabs will appear in the main part of the screen. Select Info, then choose the Server section. A drop-down screen should reveal the PHP version.


Source:PCWorld News


Share:

Your experience on this site will be improved by allowing cookies Cookie Policy