22 October 2026 · 4 min read

Analyzing the News: A Tesla and SpaceX merger

Thirteen outlets covered Elon Musk's talk of merging Tesla and SpaceX. Fíor read two of them as AI.

An empty conference table and chairs in a boardroom
Photo by SpaceX on Unsplash

Thirteen outlets covered Elon Musk's talk of merging Tesla and SpaceX. Fíor read two of them as AI.

Fíor is a natural language classifier that predicts whether a piece of text was more likely written by a person or by AI.

On SpaceX's third-quarter earnings call on 20 October, Musk said both companies had set up special committees of independent directors to explore combining them. He called it a good idea, pointing to growing overlap in AI, chips, robotics and manufacturing. He also said a deal was not certain, and that he is staying out of the talks because he has a conflict of interest. Any deal would need Tesla shareholders to approve it, which Musk said would mean at least half of all Tesla shares, not just half of those that vote.

Both companies are valued at well over a trillion dollars, so a deal would be among the largest ever attempted. Musk has hinted at it before. On Tesla's earnings call in July, he said combining companies could not be discussed on a call and would need a proper process. Now there is one.

The outlets were Reuters, AP, Bloomberg, Axios, the Guardian, TechCrunch, TIME, CBS News, Business Insider, the BBC, Quartz, Al Jazeera and The Verge. That is two wire services, business and technology outlets, American and British newspapers and broadcasters, and Al Jazeera. The Bloomberg story was scanned as published on Bloomberg Law.

The result

Eleven read as human: Reuters, AP, Bloomberg, Axios, Business Insider, the BBC, Quartz, Al Jazeera, The Verge, TechCrunch and CBS News.

Two read as AI: the Guardian and TIME.

TechCrunch and CBS News read as human, but not confidently. Every other result was clear.

This is the third story scanned with Brath, Fíor's new model. After the Google Cloud outage, where all thirteen outlets read as human, the open question was whether the new model reads almost everything that way. This story is part of the answer. It does call some coverage AI.

The Guardian has now read as AI in five of the six stories in the series where it ran its own reporting. The exception was the Google outage, where it read as human like everyone else.

Two labels that disagreed

Brath gives two answers. One is who it thinks wrote a piece. The other is how the piece reads. Until now, in this series, they had always agreed.

TIME's explainer, on why Musk wants the merger, is the first where they split. Brath judged it more likely written by AI, but said it reads like a human wrote it. That is exactly the kind of case the two labels were built to show. It could mean the text carries patterns the model links to AI while still reading with enough life to feel human. It could also mean the model is simply wrong about this one. Fíor cannot say which.

The Guardian's piece went the other way. Both labels said AI, yet Brath flagged none of the nine writing habits it looks for. Whatever made it read that way was not one of them.

Why merger coverage reads like this

This was an earnings-call story. Every outlet heard the same call, with the same quotes and the same caveats: no deal yet, Musk stepping aside, a shareholder vote to come.

The articles that went furthest beyond the call were the least decisive. TIME read as AI, and CBS News, which also explained Musk's reasoning, read as human only narrowly. It does not hold up as a rule, though. Quartz and Al Jazeera also went deep on the size of the deal and its logic, and both read clearly as human.

Some of these pieces may have had AI help, with a first draft, a rewrite or a final polish. Or this kind of coverage has become formulaic enough that a reporter and a model would write it the same way. Fíor cannot tell those apart on its own.

The committees exist because one man sits on both sides of this deal, so someone independent has to judge it. Fíor's new model has two sides of its own, and on one article they disagreed. In both cases, the final judgement belongs to people.

What Fíor is

Fíor is a machine learning classifier, currently in beta and available as a browser extension. It gives a prediction, not a verdict. It cannot see who sat at the keyboard, and nothing here is a claim about any journalist or publication. Reading as AI and being written by AI are not the same thing.

Treat it as a signal, not proof, and check the pages you read yourself.

Try the browser extension at Fíor.

Analyzing the News is a series where we run one major story through Fíor and see what comes back.

Try the extension

A free browser extension for Chrome, Edge and Firefox.