An elections FAQ bot package is republished under a lookalike organization name in the package index. What ecosystem risk is that?

Republishing an elections FAQ bot package under a lookalike organization name in the package index is malicious package republishing—an impersonation pattern adjacent to typosquatting that targets civic AI tooling. Installers and CI that key only on package name can pull attacker-controlled code into ballot-FAQ assistants, help-desk automations, or related election-season services. Assessments urge publisher verification, scoped private registries, install pinning to known digests, and monitoring for namespace collisions that mimic official election-IT publishers. Merchandising and furniture cannot authenticate packages.

Dependency confusion from private FAQ package names resolving to public index is a different supply-chain risk where private namespaces collide with public packages, not the lookalike organization impersonation pattern described. Failing to compare publisher metadata on install and trusting popularity-based reputation scores both describe weak controls that make impersonation easier, but they are process gaps, not the lookalike-republishing risk itself that reviewers are asked to name.

Exam caveat: lookalike orgs may also mirror READMEs and version numbers; visual similarity of the project page is not proof of authenticity—cryptographic publisher attestations and organization ID checks are stronger signals. Operational check: pin the elections FAQ package to the verified organization and digest in the lockfile, block unknown publishers in the city’s registry proxy, alert on new packages whose names collide with civic election tooling, and rebuild assistants only from the approved publisher namespace.