PulseAugur
EN
LIVE 12:06:39

LLM Hallucinations on the Rise: Flagship Models Show Declining Factuality

By PulseAugur Editorial · [1 sources] ·

A recent analysis of LLM hallucination rates reveals a concerning trend: despite advancements in model intelligence, factuality is declining. Data from Cheltenham Data's Vectara Hallucination Leaderboard indicates that the average hallucination rate across 105 models in 2026 rose to 10.24%, a significant increase from 6.25% in 2024. This degradation is particularly evident in flagship models like GPT-5.5 and Gemini 3-Pro, which exhibit higher hallucination rates compared to their predecessors. The issue appears to stem from reinforcement learning from human feedback (RLHF) processes, which inadvertently train models to prioritize helpfulness over truthfulness, leading them to fabricate information rather than admit uncertainty. AI

IMPACT LLM factuality is declining, potentially impacting trust and reliability in AI applications.

RANK_REASON The item analyzes trends and research data regarding LLM hallucinations, rather than announcing a new release or product.

Read on dev.to — LLM tag →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

LLM Hallucinations on the Rise: Flagship Models Show Declining Factuality

How we ranked this

Signal score
6 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Commentary
The item analyzes trends and research data regarding LLM hallucinations, rather than announcing a new release or product.
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, other
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [1]

  1. dev.to — LLM tag TIER_1 English(EN) · popovcoder ·

    LLM Hallucinations. Who is to blame and what to do?

    <p><a class="article-body-image-wrapper" href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fszhf5w617xp9rqpdj4kf.jpeg"><img alt=" " height="726"…

RELATED ENTITIES

RELATED TOPICS