AI Market Logo
BTC $43,552.88 -0.46%
ETH $2,637.32 +1.23%
BNB $312.45 +0.87%
SOL $92.40 +1.16%
XRP $0.5234 -0.32%
ADA $0.8004 +3.54%
AVAX $32.11 +1.93%
DOT $19.37 -1.45%
MATIC $0.8923 +2.67%
LINK $14.56 +0.94%
HAIA $0.1250 +2.15%
BTC $43,552.88 -0.46%
ETH $2,637.32 +1.23%
BNB $312.45 +0.87%
SOL $92.40 +1.16%
XRP $0.5234 -0.32%
ADA $0.8004 +3.54%
AVAX $32.11 +1.93%
DOT $19.37 -1.45%
MATIC $0.8923 +2.67%
LINK $14.56 +0.94%
HAIA $0.1250 +2.15%
I Watched AI Agents Try to Hack My Vibe-Coded Websit
cybersecurity

I Watched AI Agents Try to Hack My Vibe-Coded Websit

RunSybil's AI agents probe websites for vulnerabilities with machine precision, marking a new era in offensive and defensive cybersecurity.

July 31, 2025
5 min read
Will Knight

RunSybil's AI agents probe websites for vulnerabilities with machine precision, marking a new era in offensive and defensive cybersecurity.

I Watched AI Agents Try to Hack My Vibe-Coded Website

RunSybil, a startup founded by OpenAI’s first security researcher, deploys AI agents that probe websites for vulnerabilities—ushering in a new era for cybersecurity. A few weeks ago, I watched a small team of artificial intelligence agents spend roughly 10 minutes trying to hack into my brand new vibe-coded website. The AI agents, developed by startup RunSybil, worked together to probe my site to identify weak spots. An orchestrator agent, called Sybil, oversees several more specialized agents powered by a combination of custom language models and off-the-shelf APIs. Whereas conventional vulnerability scanners probe for specific known problems, Sybil operates at a higher level, using artificial intuition to figure out weaknesses. For example, it might discover that a guest user has privileged access—something a regular scanner might miss—and use this to build an attack. Ariel Herbert-Voss, CEO and cofounder of RunSybil, says increasingly capable AI models are likely to revolutionize both offensive and defensive cybersecurity. “I would argue that we're definitely on the cusp of a technology explosion in terms of capabilities that both bad and good actors can take advantage of,” Herbert-Voss told me. “Our mission is to build the next generation of offensive security testing just to help everybody keep up.” The website targeted by Sybil was one I created recently using Claude Code to help me sort through new AI research papers. The site, which I call Arxiv Slurper, consists of a backend server that accesses the Arxiv—where most AI research is posted—along with a few other resources, combing through paper abstracts for words like “novel,” “first,” “surprising,” as well as some technical terms I’m interested in. It’s a work in progress, but I was impressed with how easy it was to cobble together something potentially useful, even if I had to fix a few bugs and configuration issues by hand. A key problem with this kind of vibe-coded site, however, is that it’s hard to know what kinds of security vulnerabilities you may have introduced. So when I spoke to Herbert-Voss about Sybil, I decided to ask if it could test my new site for weaknesses. Thankfully, and only because my site is so incredibly basic, Sybil did not find any vulnerabilities. Herbert-Voss says most vulnerabilities tend to arise from more complex functionality like forms, plug-ins, and cryptographic features. We watched as the same agents tried probing a dummy ecommerce website with known vulnerabilities owned by Herbert-Voss. Sybil built a map of the application and its access points, probed for weak spots by manipulating parameters and testing edge cases, and then chained together findings, testing hypotheses, and escalating until it broke something meaningful. In this case, it did identify ways to hack the site. Unlike a human, Herbert-Voss says Sybil runs thousands of these processes in parallel, doesn’t miss details, and doesn’t stop. “The result is something that behaves like a seasoned attacker but operates with machine precision and scale,” he says. “AI-powered pen testing is a promising direction that can have significant benefits for defending systems,” says Lujo Bauer, a computer scientist at Carnegie Mellon University (CMU) who specializes in AI and computer security. Bauer recently coauthored a study with others from CMU and a researcher from AI company Anthropic that explores the promise of AI penetration testing. The researchers found that the most advanced commercial models could not perform network attacks, but they developed a system that set high-level objectives like scanning a network or infecting a host, enabling them to perform penetration tests. Sarah Guo, an investor and founder at investment firm Conviction, which is backing RunSybil, says it is rare to find people who understand both AI and cybersecurity. Guo adds that RunSybil promises to make the kind of security assessment that large companies perform periodically more widely available, and on a continuous basis. “They can do baseline penetration testing with models and tool use continuously,” she says. “So you'll always have a view of what it really looks like to be under attack.” The techniques being developed by RunSybil may become doubly necessary as attackers develop their own AI strategies. “We have to assume that attackers are already using AI to their benefit,” says Bauer of CMU. “So developing pen-testing tools that use it is both responsible and likely necessary to balance the increasing risk of attack.” Herbert-Voss seems like a good person to help here, since he was the first security researcher at OpenAI. “I built all sorts of crazy things like new prototypes of polymorphic malware, spearphishing infrastructure, reverse engineering tools,” Herbert-Voss says. “I was concerned that we didn’t have a solution for when everybody gets access to language models—including the bad guys.”
This article is an edition of Will Knight’s AI Lab newsletter.
Source: WIRED

Frequently Asked Questions (FAQ)

What are AI agents in the context of cybersecurity?

AI agents in cybersecurity are specialized programs powered by artificial intelligence that can autonomously perform tasks such as probing websites for vulnerabilities, analyzing code, and even attempting to exploit discovered weaknesses. They differ from traditional scanners by using "artificial intuition" to identify novel attack vectors rather than just known issues.

How do AI agents like RunSybil differ from traditional vulnerability scanners?

While traditional scanners look for predefined vulnerabilities, AI agents like Sybil can operate at a higher level, using intuition to discover weaknesses that might be missed by automated tools. They can chain together findings, test hypotheses, and escalate privileges to simulate a more sophisticated attacker.

What is "vibe-coded" in relation to websites?

The term "vibe-coded" likely refers to websites that are built quickly or with a focus on functionality and user experience without necessarily prioritizing robust security from the outset. These can sometimes introduce unintentional vulnerabilities that are harder to detect with standard scanning methods.

What is AI-powered penetration testing?

AI-powered penetration testing uses AI agents to simulate cyberattacks on systems to identify security vulnerabilities. This approach aims to be more scalable, thorough, and capable of discovering complex or novel security flaws compared to traditional methods.

What are the potential benefits of AI in cybersecurity?

AI can revolutionize both offensive and defensive cybersecurity by providing advanced tools for vulnerability detection, threat analysis, and automated security responses. It can help organizations keep pace with the evolving threat landscape and the increasing capabilities of malicious actors.

What are the risks associated with AI in cybersecurity?

As AI models become more capable, they can be leveraged by both defensive and offensive actors. This creates a dual-use dilemma where the same advanced capabilities can be used for both securing systems and developing more sophisticated cyberattacks.

Crypto Market AI's Take

The development of AI agents capable of sophisticated website probing and penetration testing, as demonstrated by RunSybil, represents a significant advancement in cybersecurity. This technology is directly relevant to the digital asset space, where the security of platforms, smart contracts, and user data is paramount. As AI capabilities grow, so too does the potential for both malicious actors and security professionals to leverage these tools. For those involved in cryptocurrency, understanding the evolving landscape of cyber threats and defenses is crucial. Our platform offers insights into AI-driven crypto trading strategies and the latest cryptocurrency market news, providing a comprehensive view of the intersection between AI, finance, and security.

More to Read: