Crypto protocols are facing challenges as the rise in artificial intelligence (AI) usage has led to a surge in bogus bug bounty submissions, complicating efforts to identify genuine threats. According to Cointelegraph, bug bounties, which reward ethical hackers for reporting potential vulnerabilities, are widely used in the crypto industry. While AI facilitates the analysis of extensive codebases to detect bugs, it is also prone to errors.
Barry Plunkett, co-CEO of Cosmos Labs, highlighted the impact of AI on bug bounty programs. He noted a 900% increase in submission volume from the previous year, with 20-50 reports daily, resulting in a mix of valid and invalid submissions. Kadan Stadelmann, a blockchain developer and chief technology officer at Komodo Platform, also observed a rise in bug bounty submissions and payouts across organizations. He pointed out that the decrease in report production costs due to AI might be contributing to the influx of submissions, including false positives.
In January, Daniel Stenberg, creator of the open-source data transfer tool curl, announced the termination of his bug bounty program due to the overwhelming number of AI-generated reports. HackerOne, a leading bug bounty platform, reported 85,000 valid submissions in 2025, marking a 7% increase from the previous year.
Plunkett mentioned that Cosmos Labs is adapting to the increased submissions by refining its scoring system, prioritizing trusted researchers, and collaborating with advanced triage providers. Stadelmann emphasized the importance of bug bounty programs in safeguarding decentralized systems and suggested that AI could assist in filtering out irrelevant submissions. He noted that smaller teams might struggle with the increased volume of bug bounties, highlighting the need for defensive AI systems to manage incoming reports. Stricter standards in bug bounty programs could help reduce the number of submissions, ensuring that teams can focus on genuine threats.