AI news · Sunday, October 18, 2026
Anthropic pledges 5 percent of its shares to a new benefit fund
Anthropic is trying to get ahead of the 'crisis of trust' currently hitting the AI industry by launching a new public benefit fund. The company is committing 5% of its shares to this entity immediately, with the possibility of adding another 2% for every 50% jump in company value. While it sounds like a massive philanthropy move, a 5% stake based on their May valuation is worth about $48 billion. They plan to use this fund for things like donations to nonprofits or direct aid to people hit hard by AI-driven job displacement. It is a slightly different structure than the one OpenAI set up last year, as Anthropic is keeping this as a wholly-owned LLC rather than a 501(c)(3) nonprofit, which they claim gives them more flexibility to move money around. They are also explicitly stating none of this equity will go toward political lobbying, likely trying to avoid the messy perception that AI labs are just buying influence.
Elsewhere, the consolidation in the AI market continues to look messy. Character.AI, which recently filed for bankruptcy, is essentially being absorbed by Disney. Disney isn't just buying the tech for $20 million; they are also hiring the vast majority of the startup's workforce—about 60 people—with a combined annual compensation package of roughly $60 million. It is a lifeline for the employees, though it is currently a 'stalking horse' bid, which is business-speak for 'the first offer on the table that others can still try to beat' before the final sale is locked in.
Investors are also starting to look at AI stocks with a lot more skepticism, as higher borrowing costs and fears of an infrastructure bubble have caused a sell-off in the sector. The Roundhill Neocloud ETF is down 31% this month, and companies like CoreWeave are facing scrutiny over their debt loads. Even Warren Buffett’s Berkshire Hathaway is jumping into the mix in a weird way, investing $330 million into Bending Spoons, a tech conglomerate that uses AI to acquire and aggressively 'optimize' other software companies. It is a departure from Buffett’s usual 'don't touch tech' rule, though given the amount of cash Berkshire has sitting around, it is likely more of a calculated gamble than a total pivot.
Finally, Meta managed to accidentally black out the 'No Kings' political group’s Instagram account during a nationwide voting event. Meta blamed an 'automatic trigger' that penalizes accounts for posting too frequently, proving once again that when you let algorithms run the show, they eventually target the wrong people.
The quick hits
- Anthropic is committing 5% of its shares to a new public benefit fund to support nonprofits and workers impacted by AI — a move to build trust as the company grows.
- Disney is hiring most of Character.AI's staff as part of a $20 million asset deal — a lifeline for the team after the chatbot company’s recent bankruptcy filing.
- Meta suspended a political group’s Instagram account due to an 'automated error' during a day of action — a reminder that AI moderation can accidentally silence real-world activism.
- Investors are pulling back from AI infrastructure stocks as high debt costs and fears of a compute bubble intensify — a cooling-off period for the gold rush.