Oct 23, 2025 | 8:17 AM
You might not want to rely on AI for your latest news.
A new study has found chatbots make mistakes with news content almost half the time.
The study involved 22 public media organizations in 18 countries, including C-B-C/Radio-Canada, looking at how A-I chatbots answer questions about news and current affairs.
Journalists evaluated more than three-thousand responses provided by a series chatbots like Open-A-I's Chat-G-P-T, Microsoft's Copilot, Google's Gemini and Perplexity.
They found at least one significant problem with 45% of A-I answers, serious problems with sourcing in 31% off responses and major accuracy issues in 20% of them.