Bug’s integrated businesses were bombed by AI slop

He added that there was a “third party” of “professional AI developers” who had created an “end-to-end” recording system that was “causing mass murder.”

Curl’s creator, Daniel Stenberg, wrote on a blog that the “endless decline” “has taken a toll on them mentally and taken a long time to adjust.”

Software group Nextcloud suspended its funding program in April due to a “significant increase in downgrade reports.” It said it hopes to restart the program once it finds a way to filter the data properly.

The surge in AI-powered reporting comes as Anthropic last month launched Mythos, its new cyber AI model, which it says can find software bugs faster than humans.

Companies running bug bounty programs are starting to introduce more stringent checks to combat the problem, as well as building AI assistants to test their submissions.

HackerOne, whose platform for reporting errors serves Goldman Sachs, Google, and the US Department of Defense, said it had “launched a new legal capability” this year “to help organizations manage the amount of data they have,” such as those produced by examples like Mythos.

The company said shipments rose 76 percent in the year to March. But it said the share of reports showing legitimate insecurity has remained steady over the past year at 25 percent.

HackerOne CEO Kara Sprague said in recent weeks they’ve seen an increase in “high-profile” reports that used AI. He added that the rise of AI-powered data “isn’t a strong enough excuse to say we don’t need it”, as criminals use the technology to spot more flaws.

Bugcrowd CEO Dave Gerry said events like Anthropic’s Mythos will help hunters, not replace them. “AI will help in many things but we will never replace human skills,” he said.

© 2026 The Financial Times Ltd. All rights reserved. It must not be redistributed, copied, or modified in any way.