This week offered a tidy snapshot of where artificial intelligence stands heading into the end of 2026: flush with capital, increasingly autonomous, and facing real friction as it collides with existing institutions and investor patience.

Anthropic puts a price on "public benefit"

On October 18, Anthropic announced it would create the Anthropic Public Benefit Fund, a new vehicle designed to direct equity toward charitable and economic-transition efforts as the company grows. Subject to stockholder approval, 5% of Anthropic's shares will be committed immediately, with an additional 2% added at the board's discretion for every 50% increase in the company's value above a set baseline. At Anthropic's last reported valuation, the sums are large even before any top-ups — Anthropic was last valued at $965 billion in a private funding round, meaning a 5% stake would be worth roughly $48 billion on paper.

Unlike a typical charitable foundation, the company plans to establish the fund as a wholly owned limited liability company rather than a traditional tax-exempt nonprofit. Anthropic frames the move as an alternative to the standard Silicon Valley playbook: "It is common practice for tech companies to offer escalating compensation awards to their executives. We are instead directing that equity to the public benefit." Combined with earlier founder and employee pledges, about 20% of Anthropic's shares will now be set aside for the public good, the company said. The timing — just as Anthropic reportedly readies for an IPO — has drawn scrutiny over governance and how displaced workers might actually benefit.

Banks say "not so fast" to agentic finance

While labs tout autonomous agents, the institutions managing people's money are pushing back hard. Bank of America and Citigroup have restricted third-party artificial intelligence agents from accessing customer accounts, joining JPMorgan Chase and Charles Schwab as lenders seek to protect customer funds and prevent automated transfers to higher-yielding alternatives. JPMorgan CEO Jamie Dimon was blunt on the bank's earnings call, warning the bank would limit AI agents' access to client accounts "because you don't want rogue agents and there could be thousands of them." Schwab's Rick Wurster said his firm's weekly reports show the big outside agents being stopped "in the tens and twenties" each week, specifically naming Meta's Muse. The standoff highlights a widening gap between agent developers eager to automate financial tasks and incumbents unwilling to accept the liability.

The AI trade wobbles, then rebounds

Markets reflected the tension between AI's promise and its costs. Gartner is standing by its forecast that $3.5 trillion in AI investment could be on the rocks due to profitability concerns among the top AI models. CoreWeave and other "neoclouds" sold off sharply, while Oracle bonds are sitting dangerously close to junk territory as markets cool on artificial intelligence. Yet sentiment flipped quickly: Nscale was said to be in talks to sell computing power to Anthropic in a deal that could be worth about $45 billion, helping fuel a chipmaker-led rebound.

Together, these stories capture an industry simultaneously consolidating power, distributing wealth, and testing the patience of both regulators and markets.

Sources

AI AgentsAI MarketsAnthropicArtificial IntelligenceDaily AI UpdateFintech