Tag Archives: AI

Anthropic’s business is growing fast enough that the company says it is approaching cash-flow positive territory, even as debate intensifies over how its eventual IPO should be valued. The Anthropic Q3 2026 revenue figures, reported this week at a roughly $16.8 billion quarterly run rate, mark one of the clearest signs yet that AI model providers are starting to convert massive compute spending into comparably large revenue.

IBTimes reported that Anthropic’s quarterly revenue jumped sharply from the prior period, with the company telling investors it is now at or near cash-flow positive — a milestone few AI labs burning through billions in compute costs have been able to claim. BigGo Finance’s coverage framed the number as intensifying an already heated debate over what Anthropic should be valued at whenever it does go public.

What’s driving the Anthropic Q3 2026 revenue jump

The growth is coming primarily from enterprise and developer usage of Anthropic’s Claude models through its API, alongside expanding usage of coding-focused products that have become a significant share of revenue across the AI industry this year. Anthropic has previously disclosed large compute-supply agreements, including commitments to purchase cloud capacity and chips from partners, spending that investors have been watching closely against revenue growth to judge whether the unit economics of running frontier AI models are improving.

A jump to a $16.8 billion quarterly run rate, if sustained, would represent a meaningful acceleration from earlier in the year and puts Anthropic’s revenue trajectory closer to rivals OpenAI and Google’s AI division, though direct comparisons are complicated by how each company reports usage versus contracted revenue.

Why the IPO valuation debate is heating up

Anthropic has been reported to be preparing for an eventual public listing, and earlier funding rounds have already pushed private valuations into the hundreds of billions of dollars. Time’s reporting on investor sentiment suggested some backers are tempering their valuation expectations given signs that overall AI-sector growth, while still strong, may be slowing from the breakneck pace of the past two years — even as Anthropic’s own revenue numbers this quarter moved in the opposite direction.

That tension — a single company posting accelerating revenue inside a sector some investors think is cooling overall — is likely to keep shaping how aggressively banks and investors price any eventual Anthropic offering. Reaching cash-flow positive status, if it holds up under full-year scrutiny, would also blunt one of the more common criticisms leveled at frontier AI labs: that none of them can run profitably at current compute costs.

What happens next for Anthropic

Anthropic has not set a confirmed IPO date, and companies at this stage typically wait for multiple quarters of consistent financial performance before committing to a listing timeline. The next marker to watch will be whether the cash-flow-positive status holds through the fourth quarter, and whether competitors disclose comparable figures that either support or undercut the sense that AI providers are turning a corner on sustainable unit economics.

Tamara News has also covered the regulatory side of the AI industry’s growth in its reporting on the EU AI Act’s enforcement actions against OpenAI and Anthropic and on warnings over AI data-center spending.

Frequently asked questions about Anthropic’s Q3 2026 revenue

How much revenue did Anthropic report for Q3 2026?

Anthropic’s quarterly revenue run rate reached roughly $16.8 billion, a sharp jump the company says has brought it near cash-flow positive status.

What is driving Anthropic’s revenue growth?

Growth is coming primarily from enterprise and developer usage of its Claude models via API, along with expanding demand for coding-focused AI products.

Is Anthropic profitable?

The company says it is now at or near cash-flow positive, though that is a narrower measure than full accounting profitability and has not been independently audited in public filings.

Does Anthropic have a confirmed IPO date?

No. The company has been reported to be preparing for an eventual public listing, but no confirmed date has been set.

Why are investors divided on Anthropic’s valuation?

Some backers are tempering valuation expectations amid signs that overall AI-sector growth may be slowing, even as Anthropic’s own quarterly revenue accelerated.

How does this compare to OpenAI and Google?

Direct comparisons are difficult because companies report usage and contracted revenue differently, but Anthropic’s trajectory is now seen as closer to its largest rivals than it was a year ago.

Sources

• International Business Times — Anthropic’s quarterly revenue jumps; company says it’s now cash-flow positive. https://www.ibtimes.com/anthropics-quarterly-revenue-jumps-company-says-its-now-cash-flow-positive-3808602
• BigGo Finance — Anthropic posts $16.8 billion Q3 revenue as valuation debate heats up ahead of IPO. https://finance.biggo.com/news/515564d8-131b-4f8f-a240-32f53086211b

A federal judge has sided with Apple in an early procedural fight over how much evidence it can pry out of OpenAI’s internal hiring records, giving the company’s trade-secrets case against the ChatGPT maker fresh momentum. The Apple OpenAI trade secrets lawsuit, filed in July 2026, accuses OpenAI of building its planned AI hardware around designs and know-how allegedly taken by former Apple employees who moved to the rival company.

TechCrunch reported on October 13 that the judge ordered OpenAI to turn over hiring records related to the employees at the center of the dispute, a win for Apple’s effort to establish a pattern it says goes beyond any single hire. Days later, 9to5Mac reported that Apple filed its first detailed public response to OpenAI’s countercharges, pushing back on OpenAI’s claim that Apple had improperly added new evidence to the case.

What the Apple OpenAI trade secrets case is actually about

Apple’s complaint centers on a small number of former employees who worked on in-development hardware projects before leaving for OpenAI, which has been building its own line of AI-first devices since acquiring hardware designer Jony Ive’s startup. Apple alleges those employees carried proprietary design files and technical specifications for unreleased products with them, material it says gave OpenAI an unfair shortcut in its own hardware development.

OpenAI has pushed back hard, framing the suit as an attempt by Apple to slow down a competitor through litigation rather than genuine concern over stolen material, and has separately accused Apple’s legal team of improperly expanding its claims as the case has proceeded through the fall. Neither company’s characterization of the other’s motives has been tested in court yet; both remain the parties’ own disputed accounts rather than established fact.

Why the discovery ruling matters

Getting access to OpenAI’s internal hiring records is significant because it lets Apple’s lawyers look for communications, recruiting notes or internal approvals that might show whether specific technical knowledge — rather than just general industry experience — was a factor in how OpenAI brought the former Apple employees on board. Virginia Lawyers Weekly reported this week that Apple’s filings allege at least one employee took files related to an unreleased device before departing, a claim that would be central to proving the “pattern at every level” Apple’s original complaint described.

9to5Mac also reported that Apple won early access to additional OpenAI documents and employee testimony beyond the hiring records, suggesting the discovery phase is moving faster, and more in Apple’s favor procedurally, than is typical this early in a trade-secrets dispute of this size.

What happens next in the case

With discovery now underway on hiring records and internal documents, the next milestones are depositions of the employees named in Apple’s complaint and further motions over what evidence can be used at trial. Trade-secrets cases of this scale often settle before reaching a jury, but neither company has signaled an interest in resolving the dispute privately so far, and OpenAI’s own counter-filings suggest it intends to contest the claims rather than negotiate an early exit.

For more on the AI industry’s current legal and competitive pressures, see Tamara News’s coverage of Anthropic’s warning on self-replicating AI systems and the EU AI Act’s enforcement actions against OpenAI and Anthropic.

Frequently asked questions about the Apple OpenAI trade secrets case

When did Apple file its lawsuit against OpenAI?

Apple filed the trade-secrets lawsuit in July 2026, alleging former employees took proprietary hardware design material to OpenAI.

What did the October court ruling decide?

A judge ordered OpenAI to turn over hiring records related to the employees named in Apple’s complaint, a discovery win reported by TechCrunch on October 13.

How has OpenAI responded to the allegations?

OpenAI disputes the claims and has accused Apple of improperly expanding its case with new evidence and arguments as the litigation has proceeded.

What kind of hardware is OpenAI accused of building with stolen material?

Apple’s complaint centers on unreleased AI-first hardware OpenAI has developed since acquiring designer Jony Ive’s hardware startup.

Could this case be settled before trial?

It’s possible, but neither company has signaled interest in a private settlement so far, and both sides are actively contesting evidence and discovery requests.

What happens next in the litigation?

The next steps are expected to include depositions of the named employees and further disputes over what evidence can be introduced at trial.

Sources

• TechCrunch — Judge orders OpenAI to turn over hiring records in Apple trade secrets case. https://techcrunch.com/2026/10/13/judge-orders-openai-to-turn-over-hiring-records-in-apple-trade-secrets-case/
• 9to5Mac — Here’s Apple’s official response to OpenAI’s trade secret theft allegations. https://9to5mac.com/2026/10/15/heres-apples-official-response-to-openais-trade-secret-theft-allegations

Three weeks ago, on September 3, 2026, millions of people around the world lost access to ChatGPT, Claude and Grok within the same 90-minute stretch, an incident now widely cited as the clearest evidence yet of how concentrated the AI industry’s infrastructure has become. The AI platforms simultaneous outage traced back to a single regional failure inside Microsoft Azure’s East US infrastructure, the cloud backbone that quietly underpins three of the world’s four biggest AI chatbots.

What the AI platforms simultaneous outage actually looked like

Outage reports climbed into the tens of thousands that Thursday morning as ChatGPT, Claude and Grok all became unreachable at once, knocking millions of daily workflows offline simultaneously. Downdetector spikes confirmed what users were already reporting on social media: this was not one company’s problem, but three at once. Services began recovering by 8:49 a.m. Pacific time and were fully restored by 12:38 p.m. Pacific time, meaning the worst of the disruption lasted roughly four hours from first reports to full recovery. Workers who relied on any of the three tools for coding, writing or customer support found themselves without a fallback, since many had not built any redundancy into workflows that had quietly become daily essentials.

Why one cloud region could take down three rivals

The trigger was a regional failure inside Microsoft Azure’s East US infrastructure. OpenAI, Anthropic and xAI all rely on Azure to some degree for compute capacity, meaning a single regional outage cascaded across services that most users assume compete independently of one another. Google’s Gemini, which runs on Google Cloud rather than Azure, stayed largely upright throughout the incident, with only about 500 outage reports at the peak, a stark contrast that analysts pointed to as evidence of how much resilience depends on cloud diversification.

The infrastructure risk nobody was pricing in

The incident forced a broader conversation about concentration risk in AI infrastructure that has only grown louder in the weeks since. Enterprises that built customer service tools, coding assistants or internal workflows on top of any of the three affected services discovered, some for the first time, that their AI vendor’s uptime depended on a cloud provider they had never directly evaluated. That single point of failure sits several layers beneath the branded chatbot interface most users interact with daily.

What has changed since the outage

In the weeks since, cloud diversification has become a more prominent talking point among enterprise AI buyers, with some companies asking vendors directly which cloud region underpins their service before signing contracts. None of the three affected companies has publicly detailed permanent infrastructure changes in response, and Microsoft has not released a full public post-mortem of the September 3 regional failure. The episode remains a live case study in how much of the AI boom’s visible competition sits on a surprisingly narrow band of shared cloud infrastructure.

How businesses are rethinking AI vendor risk

For companies that had built customer-facing tools on top of ChatGPT, Claude or Grok, the outage was a rare moment of shared vulnerability across otherwise competing products. Procurement teams at several enterprises have since added cloud-region questions to vendor security reviews, treating an AI provider’s underlying infrastructure as material risk information rather than an implementation detail. Some analysts have argued the incident should accelerate multi-cloud strategies across the industry, though switching an AI workload between cloud providers is rarely as simple as it sounds given how deeply model-serving infrastructure is often tied to a specific provider’s hardware.

A preview of a bigger structural question

The outage also reopened a longer-running debate about how much of the modern internet, and now the AI layer built on top of it, depends on a handful of hyperscale cloud providers. Azure, AWS and Google Cloud collectively host the overwhelming majority of the world’s AI training and inference workloads, meaning a single regional failure at any one of them carries outsized consequences by design. Whether that concentration eases over time as AI labs diversify their infrastructure, or deepens further as switching costs rise, remains one of the more consequential open questions the September 3 incident left unanswered.

Frequently Asked Questions

What happened during the AI platforms simultaneous outage?

On September 3, 2026, ChatGPT, Claude and Grok all went down within the same 90-minute window, with outage reports climbing into the tens of thousands before services fully recovered.

What caused the outage?

A regional failure inside Microsoft Azure’s East US infrastructure, which hosts three of the four biggest AI chatbots, triggered the disruption across OpenAI, Anthropic and xAI’s services simultaneously.

How long did the outage last?

Services began recovering by 8:49 a.m. Pacific time and were fully restored by 12:38 p.m. Pacific time the same day.

Was Google’s Gemini affected?

No. Gemini runs on Google Cloud rather than Microsoft Azure, and it stayed largely unaffected, with only about 500 outage reports at the incident’s peak compared with tens of thousands for the Azure-hosted services.

What does this reveal about AI infrastructure risk?

It shows how concentrated the AI industry’s cloud dependence has become; three competing chatbot services that most users assume are independent all failed together because they share the same underlying cloud region.

Related coverage on Tamara News

For more context, see our reporting on Plugin4shell vulnerability ai coding agents, Ai antitrust lawsuit and Eu ai act compliance audits.

Sources

• Axios — ChatGPT, Claude and Grok all simultaneously hit outages. https://www.axios.com/2026/09/03/chatgpt-claude-grok-outages
• 9to5Google — It’s not just you; ChatGPT, Claude, and Grok were all down in confirmed outages. https://9to5google.com/2026/09/03/chatgpt-claude-grok-outages/

Advanced Micro Devices joined an exclusive club on September 21, 2026, when its stock surged nearly 10% to an all-time high and pushed the company’s market capitalization above $1 trillion for the first time. The AMD trillion dollar valuation makes AMD only the fourth US chipmaker to cross that threshold, following Nvidia, Broadcom and Micron, and it comes on the back of a AI infrastructure buildout that shows no sign of slowing.

How the AMD trillion dollar valuation happened in a single trading session

AMD shares jumped about 9.6% to close near $614, an all-time high that pushed the company’s total market value past $1 trillion. The stock has now climbed more than 180% since the start of 2026, a run driven by accelerating demand for the data center chips that power large AI models. AMD’s most recent quarterly results showed revenue of $11.54 billion, up 50% year-over-year, with data center revenue specifically up 107% to $6.7 billion.

The deal anchoring next year’s forecast

Much of the optimism behind the rally traces to a supply agreement in which Anthropic locked in 2 gigawatts of AMD’s MI450 GPUs, a commitment large enough that AMD chief executive Lisa Su has pointed to it as a key reason she expects data center revenue to more than double in 2027. Deals of that scale give investors a clearer line of sight into future earnings than quarterly guidance alone typically provides, which analysts say explains why the market reacted so sharply.

Where AMD now sits against its rivals

AMD’s roughly $1.0 trillion valuation is about $352 billion higher than Intel’s approximately $648 billion market cap, meaning AMD is now worth about 1.5 times as much as the company it spent decades trying to catch. It remains well behind Nvidia’s valuation, but the gap with the rest of the chip sector has narrowed sharply this year as AI-related capital spending has broadened beyond a single dominant supplier.

What investors are watching next

The question now is whether AMD can convert its AI backlog into sustained earnings growth rather than a one-time re-rating driven by a single large customer deal. Analysts covering the stock are watching upcoming earnings for signs that data center momentum is broadening beyond the Anthropic agreement, and whether the broader chip rally that lifted AMD, Nvidia, Broadcom and Micron together this year can hold if AI capital spending growth slows in 2027.

The other three companies in the trillion-dollar club

Nvidia, Broadcom and Micron each reached the $1 trillion threshold before AMD, and each did so on a similar underlying story: surging demand for the specialized chips that train and run large AI models. What distinguishes AMD’s entry is the speed of the move, since much of its valuation gain has come in a comparatively short window this year rather than building gradually over several years the way Nvidia’s rise did. That speed has fueled debate among analysts over whether AMD’s valuation now fully reflects its AI opportunity or has run ahead of what near-term earnings can support.

Why the MI450 deal matters more than the headline number

The 2 gigawatt Anthropic commitment is significant not just for its size but for what it signals about AMD’s competitive standing against Nvidia, which has long dominated the market for AI training and inference chips. A deal of this scale with a leading AI lab suggests AMD’s MI450 architecture is now viewed as a credible alternative rather than a distant second choice, a shift that could pressure pricing across the sector if other AI labs follow with similar multi-year commitments. AMD executives have pointed to the agreement as validation of years of architecture investment aimed squarely at closing the gap with Nvidia’s data center dominance.

The broader chip rally AMD is riding

AMD’s milestone did not happen in isolation. Chip stocks broadly have rallied through September as AI infrastructure spending commitments from major cloud providers and AI labs have come in stronger than many analysts expected earlier in the year. That rally has lifted Nvidia, Broadcom and Micron alongside AMD, though AMD’s single-day jump was among the sharpest of the group, reflecting how directly the Anthropic deal reshaped near-term revenue expectations for the company specifically rather than the sector as a whole.

Frequently Asked Questions

When did AMD cross a trillion dollar market cap?

AMD’s market capitalization crossed $1 trillion on September 21, 2026, after its stock surged roughly 9.6% to an all-time high near $614 a share.

Is AMD the first chipmaker to reach $1 trillion?

No. AMD is the fourth US chip company to cross that threshold, following Nvidia, Broadcom and Micron.

What is driving AMD’s stock rally?

Surging AI infrastructure demand, including a deal in which Anthropic locked in 2 gigawatts of AMD’s MI450 GPUs, has anchored forecasts for AMD’s data center revenue to more than double in 2027.

How does AMD’s valuation compare with Intel’s?

AMD’s roughly $1.0 trillion valuation is about $352 billion higher than Intel’s approximately $648 billion market cap, making AMD worth about 1.5 times as much as its longtime rival.

How much has AMD stock risen in 2026?

AMD shares are up more than 180% so far in 2026, fueled by accelerating data center revenue that climbed 107% year-over-year in the most recent quarter.

Related coverage on Tamara News

For more context, see our reporting on Stock market ai trade rally, Cxmt g5 dram mass production and Global memory shortage device prices.

Sources

• Semafor — AMD reaches a $1 trillion market cap, as chip stocks drive rally. https://www.semafor.com/article/09/21/2026/amd-reaches-a-1-trillion-market-cap-as-chip-stocks-drive-rally
• The Motley Fool — AMD Just Crossed $1 Trillion. Is it Still Worth Buying Now?. https://www.fool.com/investing/2026/09/22/amd-just-crossed-usd1-trillion-is-it-still-worth-buying-now/