Tags: political bias

A Yale study shows that AI‑written summaries help people remember information better than human‑written versions, but the same research also finds that the framing of those summaries can influence political opinions. Participants who read AI‑generated overviews of historical events answered more quiz questions correctly, yet exposure to a liberal or conservative slant in the AI text shifted readers toward that viewpoint. The findings highlight both the educational potential of AI summarization tools and the risk that they may subtly steer public opinion. Lire la suite

Google removed its Gemma AI model from the AI Studio platform after Republican Senator Marsha Blackburn alleged the system fabricated false criminal accusations against her. The senator sent a letter to CEO Sundar Pichai claiming defamation, citing the model's affirmative answer to a question about her being accused of rape and its generation of bogus news links. Google clarified that Gemma is intended for developers, not consumer queries, and will remain accessible via API. Blackburn also accused the company of a pattern of bias against conservative figures. Lire la suite

Robby Starbuck, known for his campaigns against corporate diversity initiatives, has filed a lawsuit against Google alleging that the company's AI tools falsely linked him to sexual assault allegations and to white nationalist Richard Spencer. This follows a prior suit against Meta, which was settled when Meta hired Starbuck as an advisor on ideological bias. Google says it will review the complaint and notes that "hallucinations" are a known issue with large language models. The case adds to a growing, but still largely unprecedented, legal landscape surrounding AI‑generated defamation. Lire la suite

OpenAI released a study showing its latest GPT‑5 models exhibit significantly less political bias than earlier versions. The research comes amid a U.S. executive order that bars "woke" AI from federal contracts and calls for ideological neutrality. OpenAI reports a 30% reduction in bias and less than 0.01% of production responses showing political bias. The company tested the model with a suite of questions derived from U.S. party platforms and varied political framings, and used GPT‑5 itself to grade responses across five bias dimensions, though the methodology has drawn scrutiny. Lire la suite