A study from the European Broadcasting Union (EBU) and BBC finds that 45% of AI assistant answers have “at least one significant issue” such as attribution errors, hallucinations, or outdated details. Gemini was the worst platform, with 76% of its responses having significant issues. On a related note, the Indeed Hiring Lab’s GenAI Skill Transformation Index finds that 26% of jobs posted on Indeed in the past year could be “highly” transformed by AI.
Reddit has filed a lawsuit against three data-scraping companies and Perplexity for allegedly collecting and reselling Reddit posts without permission. It claims Perplexity purchased the data from the three companies despite having received a cease-and-desist warning. This is similar to a lawsuit Reddit filed against Anthropic in June.
Enterprise AI vendor Uniphore has raised a $260 million Series F at a $2.5 billion valuation. The new round brings total funding to an impressive $880.9 million. Uniphore acquired CDP ActionIQ last December.
Meta has announced agreements with USA Today, People Inc., and other news publishers to pay for access to their content and link to their articles in its AI chatbot. It’s a sharp contrast to Perplexity, which the Chicago Tribune and New York Times are separately suing for allegedly scraping content and bypassing paywalls. Despite this deal, Meta apparently still opposes compensating publishers for news links in social media posts.