Study: AI Assistants Misrepresent Nearly Half of News Responses
New EBU and BBC research finds 45% of AI news responses misrepresent facts, with errors in sourcing and accuracy across ChatGPT, Gemini, Copilot and Perplexity.
AIErrors
Read More
OpenAI’s Whisper AI: Powerful Yet Prone to Alarming Errors
OpenAI’s Whisper transcription tool, praised for its accuracy, is now under scrutiny for generating fabricated and harmful content across various industries.