There’s at Least One Job That AI Isn’t Killing

Source Domain: gizmodo.com

One of the things artificial intelligence seems particularly adept at is cybersecurity—at least if you buy the marketing pitch of models like Anthropic’s Mythos, which supposedly spotted vulnerabilities that humans failed to patch for more than 20 years. But instead of putting cybersecurity experts out of work, AI has actually created a hiring spree, according to the New York Times.

Citing headhunter companies that help place cyber engineers at major firms, the Times found that demand for these experts is so strong that recruiters have struggled to find enough bodies to fill the open roles. Glassdoor data shows that cybersecurity job listings are up 11% in the first quarter of 2026 compared to the same period in 2025, and that growth is expected to continue.

There are two ways to read the rush of cybersecurity experts: companies are so panicked by the potential hacking power of AI models—Anthropic has claimed its Mythos model is so powerful that it can’t be made publicly available because it poses unprecedented cybersecurity risks—that they are ramping up on people to fortify their defenses before the cyber apocalypse hits; or, everyone went entirely too hard on vibe coding and it’s created such a clusterfuck of bad code that companies have to hire more humans to come in and unwind the mess.

Odds are that it’s more the latter, though the preferred framing throughout the industry seems to be the former. Earlier this month, Wired reported on research that found more than 5,000 vibe-coded web apps created using popular AI software development tools that were plagued with security issues that made sensitive information trivially accessible.

We can probably safely assume that’s happening all over. Even the big tech firms have had some notable vibe-coding issues when letting AI agents step in and act as humans. For instance, Amazon reportedly had a server knocked offline because an AI agent decided to delete and recreate a database…

Source