“AI has made finding bugs cheap, but judging them is still expensive,” Jump’s Ishan Bhatt wrote in an Oct. 16 review.
Engineers still had to reproduce reported problems and check whether the proposed attacks could work. Some submissions flagged apparent weaknesses while missing protections elsewhere in the software.
The backlog peaked at 299 open reports. As reviews slowed, other researchers rediscovered bugs that had already been submitted but remained unfixed or undisclosed. About 80% of duplicates arrived more than three days after the original report.
The competition had already charged $100 per submission and limited researchers to one report a day. Jump says charging discourages researchers from submitting large numbers of speculative findings and leaving its engineers to validate them.
Its new platform also limits each main report section to 3,000 characters.
The competition uncovered serious vulnerabilities, including a scheduling error that could make Firedancer execute transactions in the wrong order and disagree with other Solana validators about the resulting network state.
Meanwhile, Jump is increasing rewards alongside the submission cost. Eligible findings receive twice the standard payout through Nov. 20, or three times for bugs specific to Alpenglow, Solana’s new voting system. Critical Alpenglow findings can qualify for up to $300,000.