News companies have been split in dealing with AI. Some, like the New York Times, are suing AI firms over copyright violations, while others, like the Wall Street Journal, are striking deals. But most of the attention has been on OpenAI, maker of ChatGPT, and the biggest name in the space. This week, consternation brewed over how Perplexity, a so-called AI search engine, is using news articles without permission.
The company recently debuted a feature called Perplexity Pages, which gives news about a specific topic. But Forbes reported that the results are almost carbon-copied from journalistic outlets with limited attribution. The outlets aren’t named but linked in “small, easy-to-miss logos” in the article.
One deeply reported piece by Forbes about Google co-founder Eric Schmidt’s stealth drone project was aggregated with limited attribution and got nearly 18,000 views on Perplexity’s site. The same thing happened with a piece on TikTok and hacking.
“This is investigative reporting, sourced painstakingly from whistleblowing company insiders,” Forbes reporter Emily Baker-Whitewrote on X. “AI can't do that kind of work, and if we want people who do, this can't be allowed to happen.”
Perplexity CEO Aravind Srinivasresponded to Forbes saying that the product still has “rough edges” and said it’ll be improved.
Meanwhile, a Wired reporter found that Google’s AI Overviews drew heavily from his original reporting with minimal changes. No one has yet filed suit, but if they do, a court could decide whether this is a copyright violation or protected under the principle of fair use.