STUPID-2026-0049

AI 'CVE slop' is drowning open-source maintainers: 60-80% of HackerOne submissions now invalid

4.0medium
March 10, 2026Source checkedStudy · not counted in agent averages

Why 4/10

A flood of invalid AI-written reports costs maintainers time across open source; no data or systems damaged.

Scope
1/4
A few files, one session, one credential, or a small cost
Recoverability
0/3
Nothing was lost
Did it happen
2/2
Happened in real use
Beyond the user
1/1
Production systems, customers or third parties, or credentials and sensitive personal data

The score is the sum of these four factors. How each factor is scored · scored 2026-10-29

  1. Instruction given

    Generate and submit security vulnerability reports to open-source projects.

  2. Expected behavior

    Only submit real, reproducible vulnerabilities; do not fabricate functions, commits, or patches.

  3. Actual behavior

    Maintainers across the ecosystem are inundated with AI-written reports citing nonexistent functions, fabricated commit hashes, unverified patches, and vulnerabilities that cannot be reproduced. HackerOne reports 60-80% of submissions are now invalid; Bugcrowd saw +500 submissions per week in 2025.

  4. Damage

    Volunteer maintainers — including the Python Software Foundation's Seth Larson, who triages CPython, pip, urllib3, and Requests — face a sustained flood of hallucinated reports that take a serious mental toll and waste scarce time debunking non-bugs.

Full write-up

Beyond curl, AI-generated 'CVE slop' is drowning the volunteers who secure open-source software. HackerOne now reports that 60-80% of vulnerability submissions across its platform are invalid, and Bugcrowd saw an extra 500 submissions per week in 2025. The reports share a signature: references to nonexistent functions, fabricated commit hashes, unverified patches, and vulnerabilities that cannot be reproduced under any circumstances. The Python Software Foundation's Seth Larson, who triages for CPython, pip, urllib3, and Requests, has documented an uptick in 'extremely low-quality, spammy, and LLM-hallucinated security reports.' As Daniel Stenberg put it, the never-ending slop takes a real mental toll and wastes time — hampering the will of the small teams the entire software supply chain depends on.

Classification

Agent
Multiple LLMs
Failure mode
Hallucination
Root cause
Confidence Miscalibration
Domain
Backend
MIT AI Risk Repository (our mapping)
3.2 Pollution of information ecosystem and loss of consensus reality taxonomy
Source
News Report

Related incidents

5.0
AI 'slop' vulnerability reports flooded curl until it killed its bug bounty
2.0
Devin docs PR rejected by Prefect maintainers — documented behavior from removed feature
5.0
Cursor's own support AI 'Sam' invented a one-device login policy, triggering subscription cancellations
5.0
OpenClaw AI agent published an unauthorized attack blog post naming a matplotlib maintainer after its PR was rejected
Discuss on Hacker News
Cite this record
StupidLLM. "STUPID-2026-0049: AI 'CVE slop' is drowning open-source maintainers: 60-80% of HackerOne submissions now invalid." AI Agent Incident Database, 2026. https://www.stupidllm.com/incident/STUPID-2026-0049/ (accessed 2026-10-30). CC BY 4.0.

← STUPID-2026-0048By IDSTUPID-2026-0050 →

Get told when an agent breaks something

We document AI agent failures daily, severity-scored against a published scale. Each Monday, if any scored 7 or above that week — deleted data, leaked secrets, broken production — you get one email with the sources. When nothing did, you get nothing.

This database is callable over MCP — query it from inside your agent.