AI News Assistants Deliver Inaccurate Information in Nearly Half of Responses, Major Study Finds

Widespread Inaccuracy in AI News Responses

Major AI assistants are routinely providing misleading and inaccurate news information across languages and territories, according to the largest-ever research study on the topic. The European Broadcasting Union-coordinated investigation, led by the BBC, evaluated over 3,000 responses from ChatGPT, Copilot, Gemini, and Perplexity against key criteria including accuracy, sourcing, and contextual understanding.

Widespread Inaccuracy in AI News Responses
Performance Disparities Among AI Platforms
Public Trust and Perception Concerns
Usage Patterns and Demographic Trends
Systemic Nature of the Problem
Industry Response and Regulatory Action

Sources indicate that 45% of all AI answers contained at least one significant issue, with more than three in ten responses showing serious sourcing problems including missing, misleading, or incorrect attributions. The report states that one in five responses contained major accuracy issues, including hallucinated details and outdated information, while 14% failed to provide sufficient context.

Performance Disparities Among AI Platforms

Analysts suggest concerning performance gaps between different AI assistants, with Google’s Gemini showing the poorest results. According to the research, Gemini had significant issues in 76% of responses—more than double the rate of other assistants. Researchers noted that Gemini’s poor performance was primarily driven by sourcing problems, particularly misattributing claims, which becomes especially problematic when the underlying information is incorrect.

The study found that while AI assistants demonstrate eagerness to answer questions, this doesn’t correlate with response quality. Across the entire dataset of 3,113 core and custom questions, only 17 responses (0.5%) resulted in refusals to answer—fewer than the 3% refusal rate found in a previous BBC survey conducted in February.

Public Trust and Perception Concerns

Researchers express particular concern about public trust in AI-generated news content. According to a separate BBC report, just over a third of UK adults say they completely trust AI to produce accurate information summaries, with this figure rising to almost half among adults under 35.

“These findings raise major concerns,” the researchers stated. “Many people assume AI summaries of news content are accurate, when they are not; and when they see errors, they blame news providers as well as AI developers—even if those mistakes are a product of the AI assistant.”, according to technological advances

Usage Patterns and Demographic Trends

The Reuters Institute’s Digital News Report 2025 indicates that 7% of online news consumers currently use AI assistants to access news, with this figure rising to 15% among users under 25. This growing adoption among younger demographics highlights the increasing importance of addressing accuracy issues in AI-generated news content., according to further reading

Peter Archer, BBC programme director for generative AI, commented: “We’re excited about AI and how it can help us bring even more value to audiences. But people must be able to trust what they read, watch and see. Despite some improvements, it’s clear that there are still significant issues with these assistants.”

Systemic Nature of the Problem

According to reports, the issues identified are not isolated incidents but represent systemic problems affecting AI assistants across borders and languages. Jean Philip De Tender, EBU media director and deputy director general, emphasized the broader implications: “This research conclusively shows that these failings are not isolated incidents. They are systemic, cross-border, and multilingual, and we believe this endangers public trust.”

De Tender added: “When people don’t know what to trust, they end up trusting nothing at all, and that can deter democratic participation.”

Industry Response and Regulatory Action

The research team has released a News Integrity in AI Assistants Toolkit aimed at developing solutions to the identified problems. Meanwhile, the EBU and its members are advocating for stricter enforcement of existing laws concerning information integrity, digital services, and media pluralism at both EU and national regulatory levels.

Archer confirmed the BBC’s willingness to collaborate with AI companies, stating: “We want these tools to succeed and are open to working with AI companies to deliver for audiences and wider society.” Researchers stress that ongoing independent monitoring of AI assistants remains essential given the rapid pace of AI development and deployment.

This coverage is based on research findings and analysis from multiple sources including the European Broadcasting Union, BBC, and Reuters Institute. For detailed methodology and complete results, readers are directed to the original research documents.