Approximate nearest neighbor algorithms trade recall for speed and return results that vary with index structure and tuning parameters. Vectors may be positioned so that they are omitted from, or injected into, the approximate result set while an exact search would rank them differently, allowing records to evade similarity-based detection or to bias generated responses. Integrity of search results is compromised.
Search Results are Manipulated via Approximate Search Error
CCC.Vector.TH06
Related Capabilities
| ID | Title | Description |
|---|---|---|
| CCC.Vector.CP02 | Vector Indexing | The service may be configured to build indexes optimized for similarity search, such as HNSW, IVF, or product-quantized structures, with user-selected algorithm and tuning parameters. |
| CCC.Vector.CP12 | Approximate or Exact Search Modes | The service can run a query either with an approximate nearest neighbor (ANN) algorithm for speed or with an exhaustive exact search for precision-critical applications. |
Related Controls
| ID | Title | Description |
|---|---|---|
| CCC.Vector.CN07 | Support Explicit ANN vs. Exact Search Configuration | Provide clients with the option to enforce exact-match (non-ANN) search where search fidelity is critical. |