Unverified50% confidenceFactExact time
Poisoning attacks designed for exact retrieval fail on the approximate HNSW indexes used in production, but the topology attack proposed in "Poisoning the Neighborhood" can exploit the structural bias of approximate indexes themselves and is harder for traditional content-filtering defenses to identify.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedPatch-level multi-vector indexes are especially vulnerable, because each vector corresponds to a local region, and an attacker can invert them one by one and reassemble them according to the geometric layout, bypassing the global information compression blurring of whole-page embeddings.68% similarUnverifiedIndirect prompt injection specifically refers to attackers carrying out attacks through external content that the model actively retrieves, making it more dangerous.67% similarUnverifiedRelying solely on content filtering and terms-of-service constraints is insufficient to completely prevent technical users from steering models toward attack scenarios.66% similarUnverifiedThe introduction of RAG expands the attack surface; attackers only need to have malicious content selected during the retrieval stage to influence outputs, without needing to poison the training data, thereby lowering the threshold for information poisoning.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1067956API
curl https://kongchang.com/api/v1/knowledge/claims/1067956MCP
get_claim(id=1067956)