LA Times‘ new AI tool labels opinionated articles, but early results raise concerns.
- LA Times now uses AI to tag articles with a “Voices” label if they take a stance.
- AI-generated bullet points summarize different perspectives, but some results are questionable.
- Journalists criticize the tool, saying it lacks editorial oversight and could harm trust.
The Los Angeles Times, under billionaire owner Patrick Soon-Shiong, has introduced an AI tool to label articles with a “Voices” tag if they express a clear stance or personal perspective. The tool also generates “Insights,” a set of bullet points offering different views on the topic. Soon-Shiong claims the move will enhance journalistic transparency and help readers navigate complex issues.
The decision has been met with resistance from the LA Times Guild. Vice Chair Matt Hamilton argues that while distinguishing news from opinion is valuable, using AI without editorial oversight could damage trust in journalism. The union is particularly concerned that AI-generated labels might misrepresent stories or introduce misleading counterpoints.
The AI system is already producing questionable results. A March 1st LA Times opinion piece about AI in historical documentaries was labeled as aligning with a “Center Left” perspective, while also suggesting that AI “democratizes historical storytelling.”
Similarly, an AI-generated insight on a February 25th article about Ku Klux Klan members in California politics noted that historical accounts sometimes framed the Klan as a reaction to societal changes rather than a hate-driven movement. While factually accurate, critics argue that presenting this as a counterpoint downplays the article’s central argument about the Klan’s lasting influence.
LA Times is not alone in struggling with AI-driven journalism. Other AI-powered news initiatives have resulted in embarrassing mistakes, such as an AI mistakenly recommending an Ottawa food bank as a tourist destination and an AI-generated list of Star Wars movies failing to follow chronological order. These incidents highlight the risks of using AI for editorial decisions without human oversight.
The use of AI in newsrooms has been growing rapidly. Bloomberg, USA Today, The Wall Street Journal, and The Washington Post all employ AI for various purposes, though few use it to assess editorial content. Patrick Soon-Shiong, who acquired the LA Times in 2018 for $500 million, has pushed for technological advancements at the paper. However, the backlash against AI-driven labeling suggests that trust in AI’s editorial judgment remains a major hurdle.
Follow CodedCuts for more updates!
