Skip to content
Submit incident
AII-1001-01383
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.

Curl ends bug bounty program due to AI-generated spam reports

curl's founder announced the end of the project's bug bounty program on HackerOne as of January 2026, citing an overwhelming flood of low-quality, AI-generated vulnerability reports that consumed significant team resources. The reports appeared technically plausible but described non-existent or meaningless vulnerabilities, prompting curl to eliminate the financial incentive structure driving the spam submissions.

26 January 2026 · curl · Fraud, scams and manipulation
26 January 2026
AII-1001-013834 sources ↓
Quick facts
Category
Fraud, scams and manipulation
Country
Sweden
Source
cnBeta.COM
Date
26 January 2026
Drafted by
Raidu Incident Detection

What happened

Daniel Stenberg, founder of curl, announced on 26 January 2026 that the project will end its security bug bounty program on HackerOne effective end of January 2026.

The decision followed an influx of low-quality, AI-generated vulnerability reports. In one week, the team received 7 HackerOne reports within 16 hours, and by late January had accumulated approximately 20 such reports.

These reports appeared logically coherent and professional but contained descriptions of vulnerabilities that did not exist or were meaningless upon human review, described as 'AI Slop'.

Reported impact

The flood of AI-generated reports forced curl's security team to expend significant effort on triage and verification work.

Stenberg indicated the volume of submissions had seriously threatened the mental health and survival of the project, given the small core maintenance team.

Index analysis

The incident reflects how generative AI systems produce hallucinations that create fake vulnerability reports with convincing but false technical descriptions.

The problem arose from financial incentive structures in bug bounty programs that reward submission volume, making them susceptible to automated spam.

Governance controls

Stenberg stated that closing the bounty was intended to remove the financial incentive that made low-effort spam submissions profitable.

Under the new policy, curl will no longer offer rewards for any vulnerability reports or assist researchers in seeking compensation from third parties.

Classification

Category
Fraud, scams and manipulation
Industry
Digital security
Country
Sweden
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Economic or property, Psychological
Harmed
Workers
Failure modes
Hallucination

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

cnBeta.COM
不堪忍受大量低质AI生成报告 Curl终结漏洞悬赏计划
26 January 2026
ai.zol.com.cn
curl终止漏洞赏金计划:AI批量生成伪漏洞报告致人力不堪重负
26 January 2026
驱动之家
Curl终结漏洞悬赏计划:不堪忍受大量低质AI生成报告!
26 January 2026
net.zhiding.cn
cURL因AI垃圾报告过载终止漏洞悬赏计划
29 January 2026