AI Model Hallucinations Fuel Slopsquatting Attacks on PyPI and npm Registries
AI models frequently hallucinate package names that do not exist in public registries, creating opportunities for supply-chain attacks known as slopsquatting. A researcher examined 139 names that five recent AI models consistently invented and checked their status on PyPI and npm as of 22 September. Seven names were already registered, with one previously hosting malicious code.
The attack works when developers copy installation commands such as pip install or npm install directly from AI responses without verification. If the name is available, an attacker can register it and include arbitrary code that runs with the developer’s privileges, exposing keys, tokens, and environment variables. This mirrors typosquatting but relies on consistent model errors rather than human typos.
Among the registered names, metro-evaluator on npm contained malicious code published in four versions on 7 December 2025 and removed five days later. npm replaced it with a security-hold placeholder. The package had been suggested by the tested models a total of ten times. Another name, css-color-stop, was registered after the list became public and contains only a 53-byte package.json file with no executable code.
Four other occupied names belong to legitimate but unrelated projects: odf, lusid, usr, and ssh-keys. The odf package on PyPI receives nearly 5,000 downloads per month and is unrelated to the OpenDocument library developers likely intended. The real weaviate client is published under weaviate-client, while the brand owner maintains an empty placeholder under the hallucinated name.
Earlier research cited in the report found hallucination rates of 19.7% across 16 models and 576,000 code snippets. Commercial models averaged 5.2% while open-source models reached 21.7%. A separate study of five newer models reported rates between 4.62% and 6.10%. One documented case showed the empty huggingface-cli package on PyPI accumulating over 30,000 downloads after an incorrect command appeared in an Alibaba repository.
AI coding agents increase the risk because they can execute installation commands autonomously. In one campaign, the name react-codeshift appeared in 47 generated files and spread to more than 237 repositories. Researchers demonstrated prompt-injection techniques that tricked agents into requesting attacker-controlled package names with success rates up to 100% on tools including Cursor, Windsurf, and GitHub Copilot.
Registry operators already block many hallucinated names through normalization and prohibition lists, leaving approximately 53 names still available for registration. The study emphasizes that developers and team leads should require human approval before AI agents install dependencies and should inspect package metadata with commands such as npm view or PyPI JSON endpoints before installation.
Related articles
OIDC Configured but GitHub Actions Still Vulnerable: Megalodon and actions-cool Attacks Expose Runner Risks
Even after widespread adoption of OIDC for temporary cloud credentials in GitHub Actions, two major attacks in May 2026 demonstrated that the approach remains insufficient on its own. Attackers compromised actions-cool/issues-helper by substituting action tags to execute malicious commits that read Runner.Worker memory. The Megalodon campaign injected malicious workflows into 5,561 public repositories, harvesting environment variables, credential files, kubeconfig, SSH keys, and OIDC tokens. OIDC eliminates long-lived AWS_ACCESS_KEY_ID secrets but leaves short-lived tokens exposed inside the runner during job execution. Four mandatory protection layers are now required: strict cloud trust policies using StringEquals on the sub claim, job-level id-token: write permissions, outbound traffic controls such as Harden-Runner, and CODEOWNERS enforcement on .github/workflows/. GitHub Environments with mandatory reviewers add an essential human-approval gate that would have disrupted the Megalodon attack path.
GitHub Repo Trap with 1774 Stars Mimics Microsoft Activation Scripts to Deliver Malware via Telegram Bot
A security researcher built a Telegram bot that daily delivers five trending GitHub repositories from trendshift.io along with README summaries. One repository named Microsoft Activation Scripts contained only a README with a reversed PowerShell command that decoded to download and execute a payload from a lookalike domain. The account was created 94 seconds before the repository and used classic one-liner installation syntax to blend with legitimate projects. Analysis of 321 trending repositories showed that simple filters such as recent account age or irm|iex patterns would incorrectly flag many legitimate projects including codex and opencode. Two independent detection rules were developed: one that normalizes and scans for obfuscated execution commands including base64, mshta, certutil and hidden window flags, and another that flags one-time owners created less than 30 days before the repository with under 10 KB of content. An additional groundedness check using the AIHOT method prevents DeepSeek from inserting unverified claims such as Ruby or Cypress into repository summaries.
GhostAction Campaign Expands to Inject Malicious GitHub Workflows Stealing AWS Keys and Tokens
The GhostAction campaign has returned with an expanded technique that targets a wider range of credentials across hundreds of repositories. On October 8, two compromised maintainer accounts injected a malicious workflow named security-audit.yml into 346 GitHub repositories, including high-profile projects such as uber/athenadriver and kitao/pyxel. The workflow disguises itself as a security audit routine and executes automatically on every code push, exfiltrating credentials via HTTP requests to attacker-controlled servers. Unlike previous waves that only harvested GitHub Actions secrets, the new version also scans the working directory and full git history for forgotten credentials in older commits. Targeted data includes AWS access keys, GitHub and GitLab tokens, Google API keys, Slack and SendGrid tokens, OpenAI and Anthropic credentials, PyPI passwords, and crates.io tokens. Since October 7, more than 500 GitHub accounts have planted similar workflows, building on earlier campaigns that affected 817 repositories in September 2025 and 772 between August and September.
Security Researcher Builds SAST Scanner for AI-Generated Code and Audits 3,800 Public Repositories
A developer released AigisSAST, a lightweight open-source static analysis tool written in pure Python with no external dependencies, specifically tuned to detect common mistakes made by AI coding assistants. The scanner was run across roughly 3,800 repositories ranging from small pet projects to popular open-source platforms. It identified thousands of potential secrets and misconfigurations, yet manual review reduced the number of genuine leaks to approximately 30 cases, mostly Telegram bot tokens, database credentials, and committed .env files. The project also examined 471 production-grade Telegram bots handling payments and VPN services, uncovering 31 repositories that exposed real credentials either in current code or in Git history. AigisSAST includes 21 detection rules, 193 regression tests, automatic remediation via the fix command, and seamless integration with GitHub Actions. The author deliberately avoided validating any discovered keys to stay within ethical research boundaries.