Calls to revoke ABC’s license erupted after a tense Oval Office exchange. President Trump lashed out at ABC News correspondent Mary Bruce for asking why he wouldn’t immediately release the Epstein files instead of waiting for Congress.
“I think you are a terrible reporter. Your news is so fake and so wrong,” Trump said, adding that the Federal Communications Commission should “take away” ABC’s license.
The clash occurred during a meeting with Saudi Crown Prince Mohammed bin Salman, where Trump accused the network of bias and called ABC a “crappy company.” The House had just voted overwhelmingly to compel the Department of Justice to release all Epstein-related records, reversing Trump’s earlier opposition. He insisted he had nothing to do with Jeffrey Epstein and claimed he expelled him from Mar-a-Lago years ago.
Meta scored a landmark victory in its antitrust battle with the Federal Trade Commission (FTC). After five years of litigation, U.S. District Judge James Boasberg ruled that the agency failed to prove Meta currently holds monopoly power in social networking. The case targeted Meta’s acquisitions of Instagram for $1 billion in 2012 and WhatsApp for $19 billion in 2014, alleging they neutralized competition. Internal emails showed Zuckerberg wrote,
“What we’re really buying is time,” referring to Instagram’s rapid growth.
But the court emphasized today’s competitive landscape, citing TikTok and YouTube as evidence of robust rivalry. The decision spares Meta from a breakup that could have reshaped its business model. Analysts warn, however, that regulatory scrutiny will persist, especially on children’s mental health and AI practices.
ChatGPT’s hallucination problem is worse than expected. A study by Deakin University found that 56% of citations generated by GPT-4o in mental health literature reviews were either fabricated or contained errors. Of 176 references analyzed, nearly 20% were completely fake, while 45% of real citations had incorrect details such as wrong page numbers or invalid DOIs. Alarmingly, 64% of fabricated replies linked to real but unrelated papers, making detection harder.
“Only 43.8% of citations were both real and accurate,” researchers wrote in JMIR Mental Health.
Accuracy varied by topic: depression citations were 94% correct, while niche areas like body dysmorphic disorder saw fabrication rates near 30%. Experts urge manual verification and stronger journal safeguards. Full findings are available in JMIR Mental Health.
