AI Hacking Panic: How Worried Should You Be About Black October?

A PHP flaw prompts outages

Cybersecurity experts are raising concerns about AI’s growing ability to find security weaknesses. This week, researchers used AI to identify a major vulnerability in PHP, a programming language used by as much as 70 percent of websites, according to The Atlantic.

Some government sites in the UK and Europe, along with sites in the Wikipedia family and the Internet Archive, briefly went offline after the flaw was found. The outages were precautionary: engineers took sites down to address the vulnerability before anyone exploited it. Experts warn that other flaws may not be found in time, especially on older websites that rely on PHP and may not receive updates.

Banks face agent concerns

AI-related security worries also extend to online banking. Axios reported Thursday that banks are detecting more unauthorized third-party AI agents operating within their systems. Account holders may have allowed those agents to act, but banks may not be able to tell whether they have permission. Executives described the situation as “horrifying,” Axios reported.

Practical steps for online safety

The National Cybersecurity Alliance recommends basic online safety practices. Be cautious with unexpected messages, check emails or texts that seem unusual, avoid suspicious links and never share personal information over the phone just because a caller sounds familiar.

A cloned voice could be used in a scam, so verify requests through another channel. Use a password manager or passkeys, and give each account a unique password of at least 16 characters. Online safety requires regular attention, not a one-time fix. Security experts call these ongoing habits “cyber hygiene.”