AI chatbots fail at accurate news, major study reveals

A major study by 22 public service media organizations, including the BBC and NPR, found that leading AI chatbots such as ChatGPT, Copilot, Gemini, and Perplexity AI misrepresent news content 45% of the time, with 31% containing serious sourcing problems and 20% major factual errors. The research, which identified Gemini as the worst performer with 72% sourcing issues, highlights systemic inaccuracies that endanger public trust and underscore the unreliability of AI for critical news consumption. Broadcasters are now calling for governments and AI companies to enforce information integrity laws and take greater responsibility for news handling, signaling potential regulatory pressures and operational risks for AI-dependent information services.

Analysis

A comprehensive study by 22 public service media organizations reveals that four prominent AI assistants—ChatGPT, Microsoft's Copilot, Google's Gemini, and Perplexity AI—misrepresent news content 45% of the time, with 31% containing serious sourcing problems and 20% major factual errors. This systemic unreliability, evidenced by instances like misidentifying German chancellors and NATO secretaries general, poses a significant threat to information integrity across languages and territories. The study's strongly negative sentiment (-0.75) highlights the severity of these findings.

The findings underscore a critical challenge for public trust, especially as 7% of online news consumers, rising to 15% for those under 25, now use AI chatbots for news. Google's Gemini exhibited the worst performance, with 72% of its responses having significant sourcing issues, while Microsoft's Copilot also showed poor results in previous and current studies, indicating persistent accuracy problems for major tech players. Per-ticker sentiment for GOOGL/GOOG was -0.8 and for MSFT was -0.6, reflecting these concerns.

In response, broadcasters and media organizations, led by the European Broadcasting Union (EBU), are advocating for governments and AI companies to take action, pressing EU and national regulators to enforce existing laws on information integrity. The "Facts In: Facts Out" campaign further demands that AI companies assume greater responsibility for news handling, signaling potential regulatory scrutiny and operational burdens for AI developers like Alphabet (GOOGL, GOOG) and Microsoft (MSFT). This regulatory pressure could significantly impact the Technology & Innovation sector.

Key Decisions for Investors

• Monitor regulatory developments closely, particularly in the EU, as potential legislation on information integrity and AI accountability could increase compliance costs and restrict deployment for AI developers like Alphabet and Microsoft.
• Evaluate the long-term reputational risk and potential user attrition for AI platforms, especially those with documented poor performance such as Gemini and Copilot, as sustained inaccuracies erode public trust and adoption.
• Assess the potential for increased operational expenses or liability for AI companies if they are compelled to implement more robust content verification mechanisms or face penalties for misinformation.