Epistemic Bias Injection: Biasing LLMs via Selective Context Retrieval

ArXi:2512.00804v2 Announce Type: replace-cross When answering user queries, LLMs often retrieve knowledge from external sources d in retrieval-augmented generation (RAG) databases. These are often populated from unvetted sources, e.g. the open web, and can contain maliciously crafted data. This paper studies attacks that can manipulate the context retrieved by LLMs from such RAG databases. Prior work on such context manipulation primarily injects false or toxic content, which can often be detected by fact-checking or linguistic analysis.