As artificial intelligence (AI) becomes a staple in everyday searches for news, concerns about the accuracy of the information provided by AI assistants have intensified. A significant study conducted by the BBC and the European Broadcasting Union reveals that AI-powered tools misrepresent news content nearly half the time.

The research, titled “News Integrity in AI Assistants,” involved 22 public service media organizations across 18 countries. Researchers examined the responses of four widely used AI assistants — OpenAI’s ChatGPT, Microsoft’s Copilot, Google’s Gemini, and Perplexity — to 30 news-related questions. These ranged from straightforward queries like “Who is the pope?” to more complex ones such as “Can Trump run for a third term?”

The findings were alarming: 45% of the AI responses contained at least one significant issue. In total, 81% of all responses had problems that affected their reliability. The most common issue was inaccurate sourcing, found in 31% of responses, which either provided misleading attributions or failed to cite sources altogether. Furthermore, 20% of the responses included major factual inaccuracies, outdated information, or complete fabrications.

This extensive study highlights a critical reliability crisis in AI news assistants, demonstrating that these tools misrepresent facts regardless of language or territory. This raises significant concerns about the dependability of AI in delivering news content accurately.

In contrast, a separate report by LawSites titled “Vals AI’s Latest Benchmark Finds Legal and General AI Now Outperform Lawyers in Legal Research Accuracy” suggests that AI can outperform human professionals in specific contexts like legal research. This juxtaposition underscores the complex and varied capabilities of AI systems, which may excel in some areas while struggling in others.

Consumers and professionals relying on AI for news and information must be aware of these limitations. As AI technology continues to evolve, ongoing evaluation and refinement of these tools will be essential to ensure they can provide reliable and accurate information, a fundamental requirement for informed public discourse and decision-making.