Valentina Palmiotti, better known in the hacking circles as Chompie, has raised alarm bells for her fellow ethical hackers. Recently crowned the most successful individual at the prestigious Pwn2Own hacking competition held in Berlin, she claims that the rise of AI tools, particularly Claude Mythos, might spell the end of her competitive days. Currently, these AI systems help her snatch up “bug bounties”—the rewards given to hackers who identify vulnerabilities in digital infrastructures before cybercriminals can exploit them. But Chompie warns that tools like Mythos are so advanced that even top-tier hackers will soon find it hard to keep pace with them.
The cyber-security landscape has been shaken to its core, and Mythos is at the center of the storm. Created by Anthropic, the tool has reportedly identified a staggering 1,600 vulnerabilities across various software applications. The creators are so concerned about its potential misuse that they have restricted its release to selected governments and cyber-security organizations. Pwn2Own, organized by the ZeroDay Initiative, invites ethical hackers worldwide to uncover flaws in specific software. This year, hackers collectively snagged nearly $1.3 million (£970,000) by uncovering 47 new hacking methods across a variety of platforms.
On the first day of the contest, Chompie demonstrated her prowess by hacking a system connected to Nvidia and snagging a cool $20,000. But it didn’t stop there. She entered what she calls “zombie hacker mode” to prepare for the next challenge. “As soon as I won the first prize, I rushed back to my hotel room to keep working on the other one. I worked from 6 PM until 6 AM without any sleep,” she revealed. Her hard work paid off as she later hacked into a Linux-based system, securing another $50,000. Footage from the event shows her exhausted yet ecstatic on stage, reflecting the intense pressure and exhilaration of competition.
Chompie described her “zombie hacker mode” as an immersive state of research and testing that lasts for hours, fueled by energy drinks and a surge of adrenaline, often while donning a black hoodie. “It’s not healthy,” she chuckled, but insisted it was a necessary part of the grind. Many champions like her have started to leverage AI tools to assist them during these intense sessions. She noted that tools such as Claude Code allow her to work faster, both in competitions and her regular role as a security researcher for IBM X-Force.
Chompie believes hackers are currently in a “sweet spot” where AI serves as a helpful ally. However, she predicts a shift is imminent with the emergence of advanced models like Claude Mythos and GPT 5.5 Cyber. “I participated in Pwn2Own this year because I thought it might be my last chance,” she confessed. “That doesn’t mean there’s no room for security research or ethical hacking, but a lot of the low-hanging fruit will start to disappear.”
She foresees a future where only the elite hackers will be able to discover new bugs and secure prizes, placing individuals like Orange Tsai—a fellow winner in Berlin—in that rarefied category. Tsai, who prefers not to reveal his real name, led his team to a staggering $375,000 (£278,000) victory by uncovering complex hacking pathways. He holds a more optimistic view about the future of human bug hunters, stating, “For me, AI feels more like a really awesome assistant that helps accelerate my research workflow.”
He acknowledges that while AI is raising the bar, human creativity and intuition will continue to play a crucial role in uncovering vulnerabilities that AI might overlook. As the landscape changes, one has to wonder: if it becomes tougher for skilled hackers to penetrate systems, what implications does that hold for cybercriminals? Research indicates that some criminals are also utilizing AI to enhance their attacks, even crafting new pathways for system intrusions, which could lead to data breaches and ransomware incidents.
However, it’s essential to note that most cyber-attacks still rely on long-established, simpler methods, such as phishing or social engineering, where attackers gain access by sending deceptive emails to unsuspecting employees. Chompie believes that AI tools will ultimately make it harder for all hackers, a development she views as beneficial for internet security. “The tide is turning against offensive hackers. I think defense stands to gain a lot from this capability,” she asserted. Yet, she emphasized that the advantages of AI for defenders in cyber security can only materialize if these tools are deployed responsibly.
The good guys need access to the most powerful tools first in order to identify and patch vulnerabilities before the bad actors can exploit them. As the landscape continues to evolve, will ethical hackers find a way to adapt and thrive, or will AI tools render them obsolete? Only time will tell…
Kaynak: Orijinal Haber