Anthropic’s IPO Filing Puts AI Economics in the Spotlight

Anthropic’s public IPO filing offers the clearest look yet at how a frontier AI company is trying to turn extraordinary demand into a profitable business—and how much computing that model requires. Elsewhere, Microsoft is offering buyouts to about 60,000 employees, Meta is pursuing a major equity financing for its AI expansion, SpaceX has secured chip financing, Apple has expanded Siri’s AI capabilities, and OpenAI has announced three new research results.

Anthropic’s IPO Filing Details Fast Growth and Large Commitments

Anthropic’s filing presents a business with sharply rising revenue alongside unusually large infrastructure obligations. Business Insider reports that the company generated $11.6 billion in second-quarter 2026 revenue, up 1,368% year over year, and recorded $974 million in quarterly operating profit. Its computing and infrastructure spending rose 300% over the same period, while other expenses increased 741% (Business Insider).

The filing also exposes the scale of the bet required to sustain that growth. Axios reports that Anthropic has nearly 2,000 customers generating at least $1 million in revenue and about 250 generating at least $10 million, while its disclosed computing commitments are described as a major risk if customer demand or model advantages weaken (Axios). The IPO will give public investors a new way to assess both the company’s margins and its dependence on continued spending.

Microsoft Offers Buyouts to About 60,000 Employees

Microsoft is offering voluntary buyouts to about 60,000 employees, according to reporting reviewed by Business Insider. The eligible group includes selected teams in the United States, Canada, India, and the United Kingdom, with some data-center, sales, and marketing employees excluded (Business Insider).

The offers vary by seniority, location, and tenure. Employees have until early December to decide whether to leave at the end of the year, and Microsoft said there are no plans for another buyout at this time. The move comes as the company continues investing heavily in artificial intelligence and data-center expansion, though the supplied evidence does not establish that AI spending alone caused the offers.

Meta Seeks Up to $95 Billion for AI Expansion

Meta is seeking to raise as much as $95 billion through common and preferred stock sales to support its artificial-intelligence expansion, Bloomberg reports (Bloomberg).

The plan includes potential investments from three major suppliers—Nvidia, AMD, and Broadcom—according to Yahoo Finance’s account of the proposal (Yahoo Finance). The evidence describes the financing as an equity raise rather than a conventional large bond sale, but the precise structure and final amount remain subject to the company’s filings and execution.

SpaceX Secures $40 Billion in GPU Financing

SpaceX has secured a binding commitment for roughly $40 billion in financing to purchase AI chips and related data-center equipment, according to a regulatory filing reported by TechCrunch (TechCrunch). Bloomberg describes the arrangement as an equipment-finance lease with Apollo Global Management (Bloomberg).

The deal is notable for its financing structure and for the scale of the planned GPU purchase. SpaceX called it the largest committed GPU financing deal, though that characterization comes from the company and has not been independently established in the supplied evidence.

Apple’s iOS 27.2 Expands Siri’s AI Reach

Apple has released iOS 27.2 and parallel updates across its platforms, with new Siri AI features at the center of the iPhone update. Ars Technica reports that Siri can now recall past conversations and draw on additional sources including voicemails and Messages attachments (Ars Technica).

The update also lets users limit which applications Siri can directly access, according to ZDNet (ZDNet). That control is an important practical detail as the assistant gains access to more personal information, although the supplied reports do not provide a fuller account of how the permission system works.

OpenAI Publishes Three More Science Results

OpenAI has announced three new results in mathematics, computer science, and physics and has placed the work on GitHub, The Verge reports (The Verge).

The company is using external experts it calls “shepherds” to guide the computer-science and physics results. For the mathematics result, OpenAI said the experts advised releasing the AI output for researchers to work through their own explanations. The announcement follows an earlier mass release of manuscripts, and the scientific significance of the new claims will depend on outside researchers’ scrutiny.

The Filing Details and Financing Plans Still Need Follow-Through

Anthropic’s IPO process will provide the next major public checkpoints on revenue, margins, customer demand, and computing obligations. Meta’s proposed financing and SpaceX’s equipment commitment also warrant attention as their final terms emerge. Apple’s Siri features are available now, while OpenAI’s results still face the slower test of external verification.