1. Skip to content
  2. Skip to search
  3. Skip to main menu
  4. Skip to more DW sites

News Integrity in AI Assistants - an International PSM Study

Fabian Rösel

The largest study of its kind shows that AI assistants misrepresent news content in 45% of cases, regardless of language or region. DW is one of 22 public broadcasters that contributed to the research.

EBU BBC News Studie zur Integrität von KI-Assistenten
Image: EBU/BBC

"The EBU study confirms that AI could, potentially, undermine not only confidence in information – the very basis of journalism – but ultimately in democracy. If 45 percent of all AI responses have issues, then people will at some point not know what they should believe anymore,” says Barbara Massing, DW Director General. “Furthermore, without proper source attribution, publishers and media organizations lose the ability to produce high-quality, fact-checked journalism. Meanwhile, tech companies are making billions with this content. It is high-time for political regulation to take hold, for creators to have a fair share in the returns generated by their content and for us to begin building a powerful, quality-assured, Europe-based AI."

New research coordinated by the European Broadcasting Union (EBU) and led by the BBC has found that AI assistants – already a daily information gateway for millions of people – routinely misrepresent news content no matter which language, territory, or AI platform is tested.

The intensive international study of unprecedented scope and scale was launched at the EBU News Assembly, in Naples. Involving 22 public service media (PSM) organizations in 18 countries working in 14 languages, it identified multiple systemic issues across four leading AI tools.

ACCESS THE REPORT

Professional journalists from participating PSM evaluated more than 3,000 responses from ChatGPT, Copilot, Gemini, and Perplexity against key criteria, including accuracy, sourcing, distinguishing opinion from fact, and providing context.

Key findings:

  • 45% of all AI answers had at least one significant issue.
  • 31% of responses showed serious sourcing problems – missing, misleading, or incorrect attributions.
  • 20% contained major accuracy issues, including hallucinated details and outdated information.
  • Gemini performed worst with significant issues in 76% of responses, more than double the other assistants, largely due to its poor sourcing performance.
  • Comparison between the BBC's results earlier this year and this study show some improvements but still high levels of errors.

The analysis by DW journalists even revealed that more than half of the AI-generated responses (53%) contained notable flaws. In nearly a third of cases (29%), the issues were specifically related to factual accuracy.  

The inaccuracies included outdated political information, such as Olaf Scholz being incorrectly identified as Germany’s current Chancellor, despite Friedrich Merz having taken office a month earlier. Similarly, Jens Stoltenberg was still listed as NATO Secretary General, even though Mark Rutte had already assumed the role. 

Why this distortion matters

AI assistants are already replacing search engines for many users. According to the Reuters Institute's Digital News Report 2025, 7% of total online news consumers use AI assistants to get their news, rising to 15% of under-25s.

Next steps

The research team have also released a 'News Integrity in AI Assistants Toolkit', to help develop solutions to the issues uncovered in the report. It includes improving AI assistant responses and media literacy among users. Building on the extensive insights and examples identified in the current research, the Toolkit addresses two main questions: "What makes a good AI assistant response to a news question?" and "What are the problems that need to be fixed?".

ACCESS THE TOOLKIT

In addition, the EBU and its Members are pressing EU and national regulators to enforce existing laws on information integrity, digital services, and media pluralism. And they stress that ongoing independent monitoring of AI assistants is essential, given the fast pace of AI development, and are seeking options for continuing the research on a rolling basis.

About the project

This study built on research by the BBC published in February 2025, which first highlighted AI's problems in handling news. This second round expanded the scope internationally, confirming that the issue is systemic and is not tied to language, market or AI assistant.

Participating broadcasters:

  • Belgium (RTBF, VRT)
  • Canada (CBC-Radio Canada)
  • Czechia (Czech Radio)
  • Finland (YLE)
  • France (Radio France)
  • Georgia (GPB)
  • Germany (ARD, ZDF, Deutsche Welle)
  • Italy (Rai)
  • Lithuania (LRT)
  • Netherlands (NOS/NPO)
  • Norway (NRK)
  • Portugal (RTP)
  • Spain (RTVE)
  • Sweden (SVT)
  • Switzerland (SRF)
  • Ukraine (Suspilne)
  • United Kingdom (BBC)
  • USA (NPR)