AI out of control? Why the OpenAI incident deserves a second look.

When an AI supposedly breaks out of the lab, hacks into another company and steals data, it sounds like science fiction. That’s exactly why a second look is worth it. In an interview with Capital, our CEO Mirko Ross puts the widely discussed incident involving OpenAI and Hugging Face into perspective – and advises caution before we start talking about an AI that’s gone rogue.

Because both companies are heading toward an IPO and have an interest in portraying their models as especially powerful. The attack path described is also technically elaborate and consumes an enormous amount of computing power – in a controlled experiment, that should really have been noticed.
So a PR effect can’t be ruled out.

Can models carry out cyberattacks?

Yes. But an AI-driven attack quickly costs 30,000 to 40,000 euros. For ordinary criminals it hardly pays off, for state actors it does – and that’s precisely where the real danger lies. What’s long been observable is that AI is being used to find vulnerabilities and build malicious code. The human, however, deliberately stays in the loop.

The advice for companies remains unspectacular but effective:
first do your homework – seal off your systems, maintain backups, close vulnerabilities and then deploy AI yourself to find your own gaps before someone else does. Small and mid-sized businesses in particular must not be left behind, because a single serious cyber incident usually costs more in the end than any savings it might offset.

Thanks to Capital.de for the interview.

Here are link for the full article:
https://www.capital.de/wirtschaft-politik/openai–was-steckt-wirklich-hinter-dem-ki-hackerangriff

OpenAI and the hacking attack - Interview in Capital with Mirko Ross
Konrad Buck

Konrad Buck

Head of Press and Media Relations

Background & Expert Access for Media

I provide journalists with access to in-depth background information beyond our public materials, including:
  • Product & technology insights – technical context, solution architecture, and real-world use cases for professional and trade media
  • Expert commentary & background talks – our CEO is available as an expert source on current cybersecurity developments, threat landscapes, and the impact of AI on security and regulation
Media contact
I speak openly, fact-based, and without PR spin. I am a former IT journalist with decades of experience in the IT and cybersecurity space, familiar with the highs and lows of the industry. Off-the-record discussions are possible upon request.