Slopsquatting
A supply chain attack where malicious actors register software package names that AI coding assistants hallucinate, exploiting developers who install the fake packages.
Slopsquatting is a type of supply chain attack where malicious actors register software package names that large language models hallucinate in their code suggestions. When a developer copies AI-generated code that references a non-existent package, they may unknowingly install malware that an attacker has pre-registered under that hallucinated name.
The name is a portmanteau of “AI slop ” (low-quality AI-generated content) and “typosquatting” (registering misspelled domain or package names).
How it works
- An AI coding assistant hallucinates a plausible but non-existent package name (for example,
securehashlibinstead ofhashlib) - An attacker monitors for commonly hallucinated package names and pre-registers them on package registries (PyPI, npm, etc.)
- A developer copies the AI-generated code without verifying the package exists
- The developer runs
pip installornpm install, pulling the attacker’s malicious package - The malicious package executes arbitrary code or establishes persistence
Origins
In 2023, security researcher Bar Lanyado demonstrated the attack vector by noting that LLMs hallucinated a package named huggingface-cli. While this sounds like the command-line tool for HuggingFace Hub, the correct installation command is pip install -U "huggingface_hub[cli]". Lanyado uploaded an empty package under the hallucinated name; in three months it received over 30,000 downloads (Wikipedia
).
The term “slopsquatting” was coined in April 2025 by Seth Larson, Developer-in-Residence at the Python Software Foundation, and popularized by Andrew Nesbitt of Ecosyste.ms (The Register, 2025 ).
Research findings
In May 2025, the academic paper “We Have a Package for You! A Comprehensive Analysis of Package Hallucinations by Code Generating LLMs” documented the scale of the problem:
- 19.7% of LLM-recommended packages did not exist
- Open-source models hallucinated far more frequently (21.7% on average) compared to proprietary models (5.2%)
- CodeLlama 7B and CodeLlama 34B hallucinated in over a third of outputs
- Researchers observed over 205,000 unique hallucinated package names across all models tested
(Spracklen et al., arXiv:2406.10279 )
Vibe coding risk
Feross Aboukhadijeh, CEO of security firm Socket, warns that developers practicing vibe coding may be particularly susceptible. Vibe coding often involves accepting AI-generated code without line-by-line review, which means hallucinated dependencies can slip through unnoticed (Wikipedia ).
Prevention
- Manually verify package names before installation
- Use dependency scanners that check packages against known registries
- Use lock files and hash verification to pin known, trusted package versions
- Never assume AI-generated code is safe without review
- Check the package’s download count and publication date on the registry before installing
Current status
As of July 2026, there has not yet been a reported case where slopsquatting has been successfully used as a cyberattack in the wild. The threat remains theoretical but credible given the research demonstrating high hallucination rates and the ease of pre-registering package names.
Sources
- Wikipedia. “Slopsquatting.” https://en.wikipedia.org/wiki/Slopsquatting
- Claburn, Thomas. “LLMs can’t stop making up software dependencies and sabotaging everything.” The Register (12 April 2025). https://www.theregister.com/2025/04/12/ai_code_suggestions_sabotage_supply_chain/
- Spracklen, Joseph et al. “We Have a Package for You! A Comprehensive Analysis of Package Hallucinations by Code Generating LLMs.” arXiv:2406.10279 (May 2025). https://arxiv.org/abs/2406.10279
- SecurityWeek. “AI Hallucinations Create a New Software Supply Chain Threat.” (April 2025). https://www.securityweek.com/ai-hallucinations-create-a-new-software-supply-chain-threat
- Cloud Security Alliance. “Slopsquatting: AI Code Hallucinations Fuel Supply Chain Attacks.” (April 2026). https://labs.cloudsecurityalliance.org/
Further reading
- What is a dependency? : Understanding npm install and why projects break — the attack surface slopsquatting exploits.
- Hallucination : The AI behavior that makes package names up in the first place.
- AI slop : Low-quality AI content — “slop” in slopsquatting comes from this term.
- Model collapse : As AI trains on AI, hallucination patterns may become more common.