Ask Onix
AI threatens future of elite hacking competitions
Valentina Palmiotti, known in cybersecurity circles as Chompie, dominated this year's Pwn2Own hacking contest in Berlin but fears her competitive days may be numbered due to advancing artificial intelligence.
Record-breaking wins at Pwn2Own
Palmiotti emerged as the most successful individual competitor at the annual event, which challenges ethical hackers to uncover vulnerabilities in widely used software. This year's contest awarded nearly $1.3 million (£970,000) to participants who exposed 47 previously unknown security flaws.
On the first day, Palmiotti demonstrated an exploit targeting an Nvidia-linked system, earning $20,000. The following day, she hacked a Linux-based system to claim an additional $50,000. Event footage captured her exhausted but triumphant on stage.
The cost of competition
Success at Pwn2Own required extreme dedication. After her initial win, Palmiotti described entering what she called "zombie hacker mode"-a grueling 12-hour sprint of research and testing fueled by energy drinks and adrenaline. "It's not healthy," she admitted, though she insisted such intensity was necessary to stay ahead.
This year, many competitors, including Palmiotti, relied on AI tools like Claude Code to accelerate their work. While these systems currently serve as assistants, she believes the landscape is about to shift dramatically.
AI's growing threat to human hackers
Palmiotti warned that advanced AI models, particularly Anthropic's Claude Mythos, could soon outpace even the most skilled ethical hackers. Mythos has already identified 1,600 vulnerabilities across hundreds of software programs, prompting Anthropic to restrict its release to select governments and cybersecurity institutions.
"I competed in Pwn2Own this year because I thought it might be my last chance. A lot of the lower-hanging fruit will start to disappear."
Valentina Palmiotti, Ethical Hacker
She predicted that only the very best hackers-like Taiwan's Orange Tsai, whose team won $375,000 (£278,000) in Berlin-would remain competitive. Tsai, however, struck a more optimistic tone, calling AI a "really awesome assistant" that frees him to explore creative ideas while the system handles repetitive tasks.
Implications for cybersecurity
While AI could raise the bar for ethical hackers, it also poses risks. Criminals are already using AI to refine attacks, though most cybercrime still relies on simpler methods like phishing. Palmiotti argued that AI's greatest potential lies in defense, provided powerful tools are released responsibly to security professionals first.
"The tide is turning against offensive hackers. Defense stands to gain a lot from this capability."
Valentina Palmiotti
For now, Palmiotti remains in what she calls a "sweet spot," where AI enhances rather than replaces human expertise. But with models like Mythos and GPT 5.5 Cyber on the horizon, the future of competitive hacking hangs in the balance.
