<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/most-rag-hallucinations-are-retrieval-failures-how-the-retrieval-brick-decides-what-the-model-can-i-8ajbd1sn7" -->

---
title: Most RAG Hallucinations Are Retrieval Failures: How the...
description: Most hallucinations in RAG systems stem from retrieval failures, not the language model inventing facts. Using the NIST Cybersecurity Framework as a concrete...
canonical: https://daily.dev/posts/most-rag-hallucinations-are-retrieval-failures-how-the-retrieval-brick-decides-what-the-model-can-i-8ajbd1sn7
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: Most RAG Hallucinations Are Retrieval Failures: How the Retrieval Brick Decides What the Model Can Invent | daily.dev
og:description: Most hallucinations in RAG systems stem from retrieval failures, not the language model inventing facts. Using the NIST Cybersecurity Framework as a concrete...
og:url: https://daily.dev/posts/most-rag-hallucinations-are-retrieval-failures-how-the-retrieval-brick-decides-what-the-model-can-i-8ajbd1sn7
og:image: https://api.daily.dev/og/posts/8aJbd1sn7.png
og:image:alt: Most RAG Hallucinations Are Retrieval Failures: How the Retrieval Brick Decides What the Model Can Invent
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Most RAG Hallucinations Are Retrieval Failures: How the Retrieval Brick Decides What the Model Can Invent

**[Towards Data Science](https://daily.dev/sources/tds)** · 12 min read · 2 upvotes · 0 comments

## Summary

Most hallucinations in RAG systems stem from retrieval failures, not the language model inventing facts. Using the NIST Cybersecurity Framework as a concrete example, the author demonstrates that naive cosine similarity ranked the correct answer (backup practices in PR.IP-4) dead last out of 55 pages, while a simple keyword match ranked it first. Three retrieval failure modes are identified: the answer not being retrieved at all, the wrong passage being retrieved, and the answer being buried among distractors. The fix is a two-part retrieval discipline: anchor on the right span using keyword matching, document structure, and embeddings as parallel signals, then scope the context tightly around that anchor. A typed answer contract at the generation layer serves as a second line of defense.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://towardsdatascience.com/most-rag-hallucinations-are-retrieval-failures-how-the-retrieval-brick-decides-what-the-model-can-invent>

## Similar posts on daily.dev

- [Your RAG Isn’t Hallucinating. Retrieval Handed It the Wrong Chunk.](https://daily.dev/posts/your-rag-isn-t-hallucinating-retrieval-handed-it-the-wrong-chunk--tw9lco7qx) · Medium · 1 upvotes · 0 comments
- [Your LLM Is Only as Good as What It Retrieves](https://daily.dev/posts/your-llm-is-only-as-good-as-what-it-retrieves-q3fmcuq2k) · Weaviate · 0 upvotes · 0 comments
- [Why production RAG systems give confident, wrong answers at scale](https://daily.dev/posts/why-production-rag-systems-give-confident-wrong-answers-at-scale-7b6euvauc) · The New Stack · 0 upvotes · 0 comments

---

Tags: [#llm](https://daily.dev/tags/llm), [#rag](https://daily.dev/tags/rag), [#vector-search](https://daily.dev/tags/vector-search)

[View this post on daily.dev](https://daily.dev/posts/most-rag-hallucinations-are-retrieval-failures-how-the-retrieval-brick-decides-what-the-model-can-i-8ajbd1sn7)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"Most RAG Hallucinations Are Retrieval Failures: How the Retrieval Brick Decides What the Model Can Invent","url":"https://daily.dev/posts/most-rag-hallucinations-are-retrieval-failures-how-the-retrieval-brick-decides-what-the-model-can-i-8ajbd1sn7","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/most-rag-hallucinations-are-retrieval-failures-how-the-retrieval-brick-decides-what-the-model-can-i-8ajbd1sn7"},"datePublished":"2026-07-15T12:20:48.102Z","dateModified":"2026-07-15T14:22:08.781Z","description":"Most hallucinations in RAG systems stem from retrieval failures, not the language model inventing facts. Using the NIST Cybersecurity Framework as a concrete...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/30b0b5dccf38917f4097bb34c935e08a?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/30b0b5dccf38917f4097bb34c935e08a?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Towards Data Science","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Towards Data Science","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/tds","url":"https://daily.dev/sources/tds"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/most-rag-hallucinations-are-retrieval-failures-how-the-retrieval-brick-decides-what-the-model-can-i-8ajbd1sn7","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,rag,vector-search","timeRequired":"PT12M"}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Towards Data Science","item":"https://daily.dev/sources/tds"},{"@type":"ListItem","position":3,"name":"Most RAG Hallucinations Are Retrieval Failures: How the Retrieval Brick Decides What the Model Can Invent"}]}
```

