You know things are getting ugly when Disney stops playing nice. The company, its ABC network and the eight local TV affiliates they own sued the FCC and Chairman Brendan Carr on Tuesday in U.S. District Court, alleging the agency is retaliating against ABC’s news coverage, The View and Jimmy Kimmel’s monologues in violation of the First Amendment.

The trigger is a licensing squeeze. Disney’s broadcast licenses aren’t scheduled to expire until 2028 through 2031, but the FCC ordered early renewal filings in April, citing an investigation into the company's DEI practices. Disney calls that pretext and has asked for a temporary restraining order and a preliminary injunction to stop the review before it starts.

The suit’s evidence is mostly public record. It points to President Trump’s repeated demands that ABC lose its licenses, and to Carr’s warning that Disney could “do this the easy way or the hard way” over Kimmel—a line delivered before the host’s brief suspension last fall. Reason frames the DEI probe as the paperwork wrapper on a speech fight.

You might recall that ABC paid $16 million in December 2024 to settle Trump’s first defamation suit, and then Paramount, Meta, X and Google struck similar deals. But now the biggest media company in the country is suing the regulator instead of writing a check. An FCC spokesperson said all broadcasters must operate in the public interest and accused Disney of a “campaign of disinformation,” per The Guardian. And Trump has been threatening this for years, even though it’s not how it works: Networks create content, but the local affiliates that air it have broadcast licenses. Disney only owns a few of ABC’s hundreds of affiliates, meaning the vast majority are owned by third parties.

And no judge has been named and no hearing date is set, so the TRO is the thing to watch—it decides whether the license review proceeds while the case grinds on. Ars Technica has the escalation in full. Everyone else who settled is presumably now reading the docket closely.

This is really dragging on, isn’t it? President Donald Trump posted on Truth Social Tuesday that “there are no talks or conversations going on, or scheduled, with the Islamic Republic of Iran.” That’s the whole diplomatic picture, roughly six months into a war that just saw oil prices rose for a third straight session as the U.S. and Iran appeared to move further away from any potential deal.

This means the naval blockade stays in force, Trump said, and the Strait of Hormuz is open with all water mines removed or detonated (but that isn’t confirmed). It’s certainly a reversal. Just over two weeks ago Trump accused Iran of being “duplicitous” for denying that negotiations were happening. Then on Monday he declined to extend a 60-day ceasefire that expired the same day.

Here’s the part that reaches your wallet. Traffic through a waterway that carried 20% of the world’s oil trade has slowed to a trickle and U.S. gasoline is averaging $4.07 a gallon—up 37% since the war began. What happens next is anyone’s guess, because nobody’s talking.

Find out if you have the best deal. Compare multiple car insurance quotes in one place and compare them to your current monthly rate. Just answer a few questions on LendingTree and get your quotes instantly.

Get Your Quotes

The companies building the most popular AI model don’t actually know what you’re doing with it. A new research project called the AI Observatory, co-led by Stanford and MIT, pulled together 85,633 conversational turns across 24,521 conversations from 5,000 consenting users and 52 models between 2023 and 2025. The takeaway? The labs’ own usage reports skew more towards work than personal matters.

The proof point is blunt. When researchers applied Anthropic’s Economic Index methodology to their dataset, it filtered out 48% of conversations—disproportionately the ones about health and relationships, harassment and hate, and sex. OpenAI’s own 2025 report found just 30% of consumer ChatGPT use was work-related. Behavior also splits by model: Anthropic for coding, Gemini for roleplay, ChatGPT for homework, Grok for news and politics, where misinformation clustered.

The caveat is real. This dataset is a rounding error next to what the labs see internally, and volunteered conversations probably undercount the sensitive stuff too. And a recent survey showed that the use of AI chatbots for mental health care among adolescents and young adults has increased 60% in just a year, a trend that survey organizers said is likely to continue.