Hackers increasingly fear AI; 'Only the best will survive'

The cybersecurity landscape is being dramatically reshaped by artificial intelligence, prompting even elite hackers to express concern about their future. Valentina Palmiotti, known as Chompie, a highly successful participant in the Pwn2Own hacking competition, stated that AI tools currently assist her in winning "bug bounties", cash rewards for identifying system vulnerabilities before malicious actors can exploit them. However, she noted that powerful AI systems like Mythos could soon make it difficult for even top hackers to compete.

Mythos, developed by Anthropic, has generated significant apprehension among cybersecurity experts. The company claims the AI has identified 1,600 vulnerabilities across hundreds of software programs. Anthropic suggests Mythos is so potentially dangerous that access might be restricted to select governments and cybersecurity institutions, fearing its misuse in attacks against any entity.

The Pwn2Own contest, organized by the Zero Day Initiative, invites ethical hackers globally to find flaws in specific products. This year, nearly $1.3 million in prizes were awarded for discovering 47 novel exploitation methods across various software, websites, and applications. All discovered vulnerabilities have been reported to the respective companies for patching.

Chompie herself utilized AI, such as Claude Code, to enhance her speed and efficiency during the competition and in her daily work as a security researcher at IBM X-Force. She described the current period as a "golden moment" where AI serves as a helpful assistant rather than a threat. However, she anticipates this will change rapidly with the advent of new models like Claude Mythos and GPT 5.5 Cyber, predicting that only the absolute best hackers will be able to discover new flaws in the future.

“

AI is more of a fantastic assistant that speeds up my research process.