In partnership with
Claude just moved into Microsoft Office. Before Microsoft did.
Claude for Word launched today in beta, letting you draft, edit, and revise documents directly from a sidebar inside Word. Edits appear as tracked changes, formatting is preserved, and it shares context with Claude for Excel and Claude for PowerPoint, meaning Claude can work across all three open documents in a single conversation.
That last part is the detail worth pausing on. Claude now has shared context across Word, PowerPoint, and Excel simultaneously. Microsoft Copilot, which Microsoft spent $13 billion building and owns through its OpenAI partnership, does not. Only 15 million Microsoft users pay for Copilot. Word alone has 450 million users. Anthropic just walked into that gap.
This is the third major Microsoft product to have Claude integrated, after Cowork and Deep Research. The question people are asking is how Anthropic managed to beat Microsoft to native AI integration inside Microsoft's own software. The answer seems to be that Microsoft needed a model good enough to actually be useful in Office, and Copilot was not cutting it.
Anthropic is trying to be everywhere and it is working.
Someone threw a Molotov cocktail at Sam Altman's house.
It happened at 3:45am. It bounced off the house and nobody got hurt. A man has been arrested. This morning, Altman posted a photo of his family and something he rarely offers: a genuine, unguarded reflection on his own failures. Conflict-averse in ways that hurt OpenAI, he wrote. Handled himself badly during the board conflict that nearly destroyed the company. A flawed person in the center of an exceptionally complex situation, trying to get a little better each year.
He also connected dots that nobody in his position usually connects out loud. The New Yorker dropped its 18-month investigation into him just days before the attack. Someone warned him the piece was making things more dangerous for him personally. He brushed it off. He is not brushing it off anymore.
His post stops short of blaming the article directly, but the implication is hard to miss. Words have consequences, and right now the rhetoric around AI and the people building it is running very hot. The most detailed investigation ever published into Altman's character drops, and days later someone throws a firebomb at his house in the middle of the night. Whatever you think of the New Yorker piece, that is a sequence of events worth sitting with. Read his full post here.
Smart starts here.
You don't have to read everything — just the right thing. 1440's daily newsletter distills the day's biggest stories from 100+ sources into one quick, 5-minute read. It's the fastest way to stay sharp, sound informed, and actually understand what's happening in the world. Join 4.5 million readers who start their day the smart way.
An Open Source Model Just Crashed the Top 3
GLM 5.1, a free open source model from a Chinese AI lab called Zai.org, just hit third place on Code Arena, the coding leaderboard where the best models in the world compete on real web development tasks. It jumped over GPT-5.4, Gemini 3.1, and is now sitting right alongside Claude Sonnet 4.6. It has never been this close before.
Open source models are free to download and run by anyone, no subscription, no API costs. The fact that one just cracked the top 3 against the most expensive closed models in the world is the kind of thing that makes people in the industry pay attention. The gap that big labs have been counting on is closing faster than expected.
A year ago, open source models were a curiosity. Now they are competing for the podium.