← Back to Blog

This Week in AI — Episode 19: Anthropic Files to Go Public, Big Tech's Spending Test, and the Web Starts Logging In

Covering October 24–30, 2026: Anthropic's S-1 goes public, Wall Street grades the AI build-out, a trillion-parameter open model arrives, AI search squeezes the open web, and agents reach the whole office.

Editorial illustration of this week's AI developments: a stock exchange listing, Big Tech earnings charts, an open-weight AI model, and a login wall closing off part of the open web.

The week of October 24–30, 2026 was the week AI met the public markets. An AI lab published its finances for anyone to read, Wall Street judged whether hundreds of billions in data center spending is paying off, and one of the internet's largest communities explained, in hard numbers, what AI-powered search is doing to its traffic.

Quick Answer

This Week in AI — Episode 19 covers five stories from October 24–30, 2026: Anthropic filed to list on Nasdaq under the ticker ANT, revealing $16.3 billion in first-half revenue. Alphabet, Microsoft, Meta and Amazon reported earnings that pulled in different directions, with Alphabet falling about 12% despite Google Cloud more than doubling. Mistral released open weights for its 1-trillion-parameter Large 4 model. Reddit said Google's AI search features are costing it over a million visits a week and will test a logged-in-only site. And Uber is giving more than 30,000 employees OpenAI's ChatGPT Work agent.

For most of the AI boom, the biggest questions were answered behind closed doors — in private funding rounds, lab benchmarks and partnership announcements. This week, many of them moved into the open. Here's what happened and why it matters.

$16.3B
Anthropic's first-half 2026 revenue
−12%
Alphabet's one-day share drop
1T
Parameters in Mistral Large 4
1M+
Weekly search visits Reddit lost in October

1 What Did Anthropic Reveal in Its IPO Filing?

Anthropic, the company behind Claude, filed its S-1 registration with the U.S. Securities and Exchange Commission on Monday, October 26, formally starting the process of listing on Nasdaq under the ticker ANT.

The filing gives the first official look at the company's finances. Anthropic reported $16.3 billion in revenue for the first half of 2026, compared with about $1.16 billion in the same period a year earlier — growth of roughly 1,300%. It also describes heavy, ongoing spending on computing power and makes no promise of sustained profitability.

What stood out to many readers was the risk section. Close to a quarter of the 94-page document is devoted to risk factors, including competition with OpenAI and Google, dependence on cloud computing partners, changing AI regulation, and the possibility of a serious AI-related incident that could be difficult to insure.

The offering is led by Goldman Sachs and Morgan Stanley, with both Anthropic and existing shareholders selling shares. The share count, price and valuation have not been disclosed yet. Anthropic's last private round in May valued it at $965 billion, and the Wall Street Journal reports the company is targeting the week of November 16 for its listing.

Why This Matters

Until now, the economics of frontier AI labs were mostly visible through leaks and investor chatter. A public filing puts real revenue, real costs and real risks on the record — and gives every other AI company a benchmark to be measured against.

💡

What to watch: The price range and final valuation, expected in an amended filing closer to the listing date; how public investors weigh Anthropic's growth against its compute costs; and whether OpenAI speeds up its own path to the public markets.

2 Is Big Tech's AI Spending Paying Off?

Four of the largest AI spenders reported quarterly results on October 28–29, and the market's reaction showed just how divided investors are about the AI build-out.

Alphabet · −12%
Cloud Doubled, Shares Still Fell

Google Cloud revenue grew 103%, yet Alphabet shares dropped about 12% on October 29. Investors focused on slowing search advertising growth, an $8.7 billion loss on equity investments that pulled earnings below expectations, and a second straight quarter of negative free cash flow. Capital spending of $51.4 billion also came in below the roughly $55 billion analysts expected.

Microsoft · Azure +48%
An Earnings Beat, but Lighter Spending

Microsoft beat earnings forecasts and grew Azure 48%. Its capital spending of $45.2 billion was below the $50.7 billion expected, which the company attributed to delays in capacity from third-party data center providers.

Meta · +3%
The Only One Investors Rewarded

Meta rose about 3% after revenue grew 26% to $64.6 billion. It also raised the low end of its 2026 spending plan to $135–145 billion.

Amazon · AWS +39%
Fastest Cloud Growth in Years, Heaviest Spending

AWS revenue rose 39% to $46.0 billion, its fastest growth in nearly five years. But Amazon spent $77.3 billion on property and equipment in the quarter, trailing-twelve-month free cash flow swung to an outflow of $42.4 billion, and earnings per share of $1.89 came in below estimates. Shares fell in after-hours trading.

The lighter-than-expected spending from Alphabet and Microsoft rippled out to companies that sell AI computing capacity: Oracle and Nebius fell about 3%, and Fastly dropped 6%.

Key Insight

AI demand is real — cloud growth of 39% to 103% is enormous for businesses this size. But investors are no longer rewarding spending on its own. The question has shifted from "how fast are you growing?" to "when does the cash come back?"

3 What Is Mistral Large 4 and Why Does It Matter?

French AI company Mistral released the open weights for Mistral Large 4 on October 29, about three weeks after a public preview. Anyone can now download the model from Hugging Face and run it on their own infrastructure.

Large 4 has 1 trillion total parameters but uses a mixture-of-experts design that activates about 49 billion per token, which keeps it far more efficient to run than its size suggests. It supports a 1-million-token context window, handles both text and images, and works across more than 160 languages, including every official EU language.

Mistral has tuned the model for software development, finance and spreadsheet work, legal and business workflows, and security operations teams. API pricing is $0.70 per million input tokens and $2.10 per million output tokens — roughly half the preview price.

💡

The licensing catch: The model is free for internal use, but companies that offer it as a hosted service to others need a separate commercial agreement with Mistral. Most published benchmark results also come from Mistral itself, so independent testing over the coming weeks will matter.

Why This Matters

Open models at this scale give companies and governments — especially in Europe — a serious option they can run and control themselves, rather than relying entirely on a handful of large cloud providers.

In its October 29 earnings report, Reddit said traffic from search engines fell meaningfully in the third quarter — and that in October it lost more than one million search referrals per week on average. CEO Steve Huffman pointed directly at Google's AI Overviews and AI Mode, which answer questions without sending people on to the original source.

Web search is evolving from a gateway to the web to becoming an AI-powered destination itself.

— Steve Huffman, Reddit CEO, Q3 2026 earnings call

Reddit's business is still growing — revenue rose 53% to $895 million — but U.S. daily users declined for the first time since its 2024 IPO, and the company warned total users could fall by a low double-digit percentage next quarter. Its shares dropped about 25% on October 30.

Reddit's response is significant: it will begin testing a logged-in-only experience in select markets outside the U.S. this quarter. Logged-in users already drive more than 80% of its revenue, and a login wall also makes it harder for AI companies to collect Reddit content without a licensing deal.

Why This Matters

If large sites start closing their doors to protect traffic and data, the open web that AI models learned from — and that search used to connect us to — could look very different in a few years. This is the shift from SEO to AEO happening in real time: being the answer now matters more than being the link. Read our SEO vs AEO guide →

5 How Is Uber Rolling Out AI Agents to 30,000 Employees?

Uber is expanding OpenAI's ChatGPT Work agent to more than 30,000 employees, following a pilot with 1,000 staff. OpenAI announced the rollout on October 22.

ChatGPT Work handles multi-step office tasks such as preparing meeting briefs, building slide decks and analyzing data, and it connects to Gmail, Google Drive, Google Calendar and Slack. It runs on GPT-6.1 Sol by default — a model OpenAI launched last month at about a fifth of the price of its top model. Uber's engineers will continue using Codex for coding work.

As AI transforms how we work, we want to give everyone the tools to innovate, solve problems and help shape Uber's future.

— Praveen Neppalli Naga, Uber CTO

The timing is notable. Earlier this year, Uber's CTO said the era of spending freely on AI was ending, and the company reportedly used up its full-year budget for one AI coding tool by April. Choosing a lower-cost default model for a company-wide rollout reflects that shift.

Key Insight

Enterprise AI is moving from small pilots to whole-company deployments — and cost per task is becoming just as important as capability.

◆ ◆ ◆

What This Means for You

This week, AI's biggest questions moved into public view. Anthropic's filing shows how fast AI revenue can grow — and how much it costs to get there. Big Tech earnings showed that investors are no longer rewarding spending on its own; they want to see cash flow follow. Mistral showed that open models are keeping pace at the very top end. And Reddit's numbers made concrete something many website owners have felt for months: AI answers are changing who gets the visit. For anyone building a business, a career or a website, the value is shifting toward whoever owns the relationship with the user — whether that's an AI assistant, a platform with logged-in members, or a company that has put AI tools in every employee's hands. Stay Curious. Explore What's Next. Grow Together.

Frequently Asked Questions

Anthropic filed its S-1 registration with the U.S. SEC on October 26, 2026, to list on Nasdaq under the ticker ANT. The filing reported $16.3 billion in revenue for the first half of 2026. The share price and listing date had not been announced as of October 30, though reports say it is targeting the week of November 16.

Alphabet shares fell about 12% on October 29, 2026, despite Google Cloud revenue growing 103%. Investors reacted to slowing search ad growth, an $8.7 billion equity investment loss that lowered earnings, and a second straight quarter of negative free cash flow.

Mistral Large 4 is a 1-trillion-parameter open-weight AI model from French company Mistral, with about 49 billion parameters active per token and a 1-million-token context window. Its open weights were released on October 29, 2026, and are free for internal use.

Reddit said Google's AI Overviews and AI Mode cost it more than one million search referrals per week in October 2026. It plans to test a logged-in-only experience in select non-U.S. markets in Q4 2026, since logged-in users drive over 80% of its revenue.

ChatGPT Work is OpenAI's AI agent for multi-step office tasks like meeting briefs, slides and data analysis. Uber is rolling it out to more than 30,000 employees after a 1,000-person pilot, using GPT-6.1 Sol as the default model.

Topics: Anthropic IPO Big Tech Earnings AI Spending Mistral Large 4 Open-Weight AI Reddit AI Search AI Agents