Skip to content
Submit incident
AII-1000-44157
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.

AI-generated misinformation in government opinion reports, Henan

The Democratic League of Henan reportedly found that AI-generated content with fabricated cases and data had been submitted to its public opinion reporting system, affecting decision-making information. The organization issued oversight requirements to prevent AI-generated misinformation in future submissions.

July 2026 Β· Democratic League member organizations across Henan Β· Reliability and hallucination
July 2026
AII-1000-441572 sources ↓
Quick facts
Category
Reliability and hallucination
Country
China
Source
ε‡€ε‡°η½‘οΌˆε‡€ε‡°ζ–°εͺ’体)
Date
July 2026
Drafted by
Raidu Incident Detection

What happened

On 15 August 2026, the Democratic League Henan Provincial Committee issued a report on public opinion information submitted in July 2026, noting that some submissions contained clear signs of AI generation.

The committee received 261 public opinion reports from member organizations across Henan that month, with 97 adopted at a 37.2% adoption rate. The report identified that some manuscripts contained fabricated cases and data due to 'AI hallucination' and were submitted without proper verification of factual accuracy.

Reported impact

The use of AI-generated content without verification seriously compromised the credibility of information and the quality of submissions submitted to government decision-makers.

Index analysis

The Democratic League identified AI creation abuse and inadequate review oversight as one of four major problems in the submission process. Some manuscripts showed obvious AI generation traces and contained fabricated or false information, with responsible units failing to strictly verify the authenticity of content before submission.

Governance controls

The committee emphasized the need to establish strict manuscript review mechanisms, with each case and data point verified one by one, prioritizing frontline research results and authoritative industry data. If news materials are cited, they must come from official mainstream media, and the submission of unverified data, false cases, and AI-fabricated content must be strictly prevented.

The committee required that all submitted public opinion information must strictly include four core sections: summary, introduction, problem analysis, and policy suggestions, with complete format, clear structure, and rigorous logic. Manuscript sources must be traceable and verifiable.

Classification

Category
Reliability and hallucination
Industry
Government, security and defence
Country
China
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Public interest, Reputational
Harmed
Government
Failure modes
Hallucination, Fabricated citation
AI systems
Unspecified generative AI systems

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

ε‡€ε‡°η½‘οΌˆε‡€ε‡°ζ–°εͺ’体)
ζ°‘η›Ÿζ²³ε—ηœε§”δΌšι€šζŠ₯οΌšζ”Άεˆ°ηš„δΈ€δΊ›η€Ύζƒ…ζ°‘ζ„δΏ‘ζ―ε­˜εœ¨ζ˜Žζ˜ΎAIη”Ÿζˆη—•θΏΉ
16 August 2026
ε‡€ε‡°η½‘οΌˆε‡€ε‡°ζ–°εͺ’体)
ζ°‘η›Ÿζ²³ε—ηœε§”δΌšι€šζŠ₯οΌšζ”Άεˆ°ηš„δΈ€δΊ›η€Ύζƒ…ζ°‘ζ„δΏ‘ζ―ε­˜εœ¨ζ˜Žζ˜ΎAIη”Ÿζˆη—•θΏΉ
17 August 2026