BugHunt reports 115% growth in active researchers and increased payments, thanks to AI

An internal survey by BugHunt reveals that the number of active researchers on its Brazilian platform increased by 115% between September 2025 and September 2026, compared to the previous year. The company attributes this growth partly to artificial intelligence, which has simplified the process of onboarding new participants.

Growth in the researcher community

The platform has over 27,000 registered users, but only those who submit at least one vulnerability report are considered active. During this period, new registrations rose by 44%, and the proportion of first-time reporters increased from 64% to 77.8% of all active researchers. Meanwhile, the concentration of submissions among the top 10% most active researchers decreased from 69.1% to 63.4%.

The role of artificial intelligence

The bug bounty platform also saw an increase in payments, with the average reward rising by 17% and the total payout growing by 22.6%. According to the founders, automation has led to a higher volume of work, requiring skilled professionals to validate findings and oversee their resolution.

“We’ve always had a large community, but some members struggled to get started. They lacked the technical skills to turn their curiosity into a report. AI has made this process much easier,” says Caio Telles, co-founder of BugHunt.

“AI generates more work, and that work requires people who can distinguish genuine findings from noise and ensure issues are fully resolved,” adds Bruno Telles, another co-founder.

Challenges with invalid reports

Since March 2026, BugHunt has noticed a surge in submissions consistent with automated report generation. Monthly submissions were two to four times higher than the previous year’s average. As a result, reports without technical validity nearly doubled, and spam submissions more than tripled. The company links this trend to automation but does not specify how much of the total was generated by AI.

Improved operational efficiency

Despite the increase in submissions, the median time to close a report dropped from 15.2 to 3.7 days. Among approved reports, the share classified as critical or high severity rose from 58.2% to 68.5%, and the proportion of reports with correctly filled technical classifications increased from 45.2% to 61.0%.

Conclusion and outlook

The data suggests that AI is making bug bounty programs more accessible and efficient, but experienced teams remain essential for evaluating and fixing vulnerabilities. BugHunt emphasizes that technology does not replace researchers but expands participation and requires stronger screening capabilities.