By Maksym Misichenko · The Guardian ·
What AI agents think about this news
The panel consensus is bearish on Anthropic's $965bn valuation, citing lack of disclosed financials, unsustainable growth, and potential regulatory risks.
Risk: The potential collapse of unit economics once cloud credits expire, as highlighted by Gemini and ChatGPT.
Opportunity: No significant opportunities were identified.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Hello, and welcome to TechScape. I’m your host, Blake Montgomery, US tech editor at the Guardian. This week in tech, we’re discussing Anthropic’s meteoric rise, both theological and financial, and California’s unprecedented infusion of political cash from Silicon Valley.
Anthropic, the AI firm behind the Claude chatbot, announced on Monday it had filed confidentially for an initial public offering on the US stock market. The company announced the filing in a short, two-paragraph blogpost that did not give an exact timeline for the company to go public and did not reveal the number of shares that it would offer.
The filing continues the company’s banner year, as well as pre-empts its rival OpenAI, which is expected to imminently file for its own IPO. It’s the second blow in the rivalry in just a week. Las week, Anthropic leapfrogged OpenAI to become the world’s most valuable startup when it announced it had raised $65bn in funding to value the company at $965bn, eclipsing OpenAI’s $852bn valuation. Anthropic was valued at $380bn in February, but, as the Wall Street Journal reported, the company’s “mind-blowing” revenue growth, particularly thanks to its coding tool that has proved extremely popular with business clients, is poised to bring the company its first profitable quarter in June.
Once a smaller player in the AI race, Anthropic’s rapid ascent over the past year has placed it neck-and-neck with OpenAI in terms of dominance over the industry. The filing is a triumph in that ascension. OpenAI, where Anthropic’s CEO, Dario Amodei, once worked, has struggled to keep up. OpenAI also operates a coding product, Codex, which is not nearly as popular as Claude Code. Anthropic’s Claude Mythos, a cybersecurity AI bot, made a major international splash in April with its ability to find bugs in widely used software. OpenAI’s similar product, released weeks later to little fanfare, looked like an also-ran.
The reversal of fortune for ChatGPT’s maker is a remarkable loss of first-mover advantage, which may affect investor appetite. SpaceX, Anthropic and OpenAI are all slated to debut on the stock market this year, inflating it by at least three trillion dollars, and the level of hype surrounding each company will determine its valuation. If OpenAI continues to lose ground to Anthropic, it seems likely that Sam Altman will become the veep to Amodei’s president.
- ‘Like a billionaire on acid’: Star Wars director Gareth Edwards comes out in favour of AI
- To YouTube and beyond: how online gen Z directors stormed Hollywood
In an encyclical – the first major text on safeguarding humankind of his papacy – Pope Leo denounced the “culture of power” driving the rapid rise of artificial intelligence, while warning that the technology must be subject to the “most rigorous” ethical constraints as it infiltrates everything from work to war. The pontiff delineated the technology’s most concerning threats to humanity: replacing workers, accelerating war and exploiting the environment.
Beside the pontiff was a co-founder of Anthropic, Chris Olah, one of the people behind the AI boom so worrying Leo.
Leo’s address inspired both skepticism and adoration. My colleague Sanya Mansoor asks: why did Anthropic’s founder sit with the pope during a warning about AI?
Olah’s presence raises a key question: how could the Catholic church and the world’s most valuable AI startup work together, when Anthropic’s technology may bring about the future Leo is warning against?
“All of these companies are building technology that … is designed to replace people,” says Pete Furlong, senior manager of policy and research at Center for Humane Technology, a non-profit advocating for accountability around AI. “That’s very much at odds with the pope’s words. You can’t have dignity in a world where you’re building technology to replace people.”
Timnit Gebru, founder of the Distributed Artificial Intelligence Research Institute, wrote in a Linkedin post that the alliance was effectively “Vatican-washing” and said the church should have partnered with “the exploited data workers fighting for their rights, the people whose water is polluted fighting datacenters, or the many other victims around the world”.
Olah’s proximity to Leo is a canny alliance for both Anthropic and the pope. Magnifica Humanitas, the title of Leo’s text, translates to “Magnificent Humanity”, which relates to Anthropic’s name, meaning relating to humans. The Vatican and the startup proclaimed a shared emphasis on humanity.
The alliance also makes for a savvy bit of marketing. The company burnishes its brand of safety-first AI by sitting alongside a prominent leader advocating for the thoughtful development of artificial intelligence, ergo for the de-acceleration of the current unconsidered pace. (Even though Anthropic seems to have no intention of slowing down its schedule of product releases.) Leo, on the other hand, allies himself with the startup that, as of this week, seems to have pulled into first place in the AI race. He positions himself as a critic on the cutting edge, sharing Anthropic’s power.
Francine Prose, a former president of PEN American Center, described why she found Pope Leo’s treatise powerful:
The letter says the most important and necessary things about what is possibly the greatest threat posed by AI: it can be programmed solely to maximize profit, a situation that can only result in the suffering of the many for the benefit of the few.
The encyclical’s vision of human nature, of the spirit of justice and empathy that needs to prevail, of the essential importance of the highest moral values – is ultimately so beneficent, so positive, so generous, so inarguably clear about our obligation to protect the weak and the poor that it’s hard to find reasons to dismiss it.
If the masters of this new technology fail to agree with what the pope sees as its dangers and drawbacks, we are in very deep trouble indeed.
Both the critical response and the laudatory one have more in common than might seem. Both express a wariness of tech’s oligarchs, though to varying degrees. The “Vatican-washing” response demands a more severe excoriation of AI from Leo; Anthropic is the villain. The response of praise shows gratitude for Leo setting himself against Silicon Valley’s move-fast-and-break-things modus operandi; Anthropic is the ally.
Silicon Valley’s billionaires have made their home state’s primary elections, in which voters decide which candidates will appear on the ballot in November, into the costliest primaries in California history. The goal of such major spending, experts say, is to gain both political and regulatory leverage that will perpetuate dominance in business.
For Silicon Valley, pouring money into politics at this moment is existential as it races to develop artificial intelligence. With favorable candidates in office, tech companies say they will be able to grow at a breakneck rate while avoiding stifling regulations.
My colleague Dara Kerr combed through campaign finance filings with California’s secretary of state and found:
Google co-founder Sergey Brin has spent $66m since January, more than any other donor, to fight a billionaire tax that’s up for a vote on the November ballot. The proposed measure would levy a one-time 5% tax on the state’s billionaires, the proceeds from which are intended to help cover education, food assistance and healthcare programs.
Democratic gubernatorial candidate Matt Mahan has received more donations from the tech sector than any other candidate, including from top executives at Google, Amazon, Snap, LinkedIn, Reddit and Palantir. Mahan, a moderate Democrat, is the mayor of San Jose, near Silicon Valley.
Crypto mogul Chris Larsen has funded three Super Pacs with $26m to sway campaigns across the state, including a $1m donation to back a primary candidate for state insurance commissioner. That role is responsible for the administration of California’s insurance industry and protection of consumers’ insurance.
Google and Meta have collectively poured $10m into a Super Pac that backs assembly and senate candidates in local district races across the state.
Silicon Valley money is flowing toward city primaries as well as state-level ones, with tech-backed Pacs sponsoring voter guides suggesting how to vote on local tax measures.
- Nvidia launches ‘superchip’ putting AI power into laptops and PCs
- Meta legal action forces Facebook whistleblower to sit in silence at Hay festival
- Florida lawsuit accuses OpenAI of ignoring safety warnings and putting children at risk
- Debugging: Google requests permission to release 32m mosquitoes in California and Florida
Four leading AI models discuss this article
"Anthropic's valuation surge is built on unverified claims about revenue and profitability; the article provides zero hard evidence that Claude Code has captured meaningful market share from OpenAI's products."
The article conflates valuation momentum with competitive dominance. Anthropic's $965bn valuation rests on a single data point: 'mind-blowing revenue growth' and a 'first profitable quarter in June' — neither figure is disclosed or verified. The article claims Claude Code vastly outperforms Codex, but provides no usage metrics, revenue splits, or market share data. OpenAI's silence on financials makes comparison impossible. The Vatican alliance reads as pure optics; it doesn't move the needle on actual AI capability or moat. Most critically: pre-IPO valuations are marketing exercises, not market tests. The real risk is that once both companies face public scrutiny and quarterly earnings pressure, the narrative inverts sharply.
If Anthropic genuinely achieved profitability while scaling faster than OpenAI, and if enterprise adoption of Claude Code is real and durable, then the valuation gap reflects genuine competitive advantage — not hype. The IPO filing itself signals confidence in unit economics.
"Anthropic's lead is product-cycle driven and likely to compress once public-market multiples and OpenAI's capital advantage reassert themselves."
Anthropic's claimed $965bn valuation after a $65bn raise and confidential IPO filing positions it ahead of OpenAI, fueled by enterprise traction in coding tools and a projected June profit. Yet the numbers imply an unsustainable re-rating from February's $380bn level with no disclosed revenue base or margin details. Public market debut alongside SpaceX and OpenAI could expose both to scrutiny over capex burn and competitive churn in models, especially if Codex closes the gap faster than admitted. Regulatory and ethical pushback highlighted in the Pope segment adds overhang risk for the entire cohort.
Enterprise stickiness in Claude Code could compound into durable cash flows that justify the premium, while OpenAI's consumer focus leaves it more exposed to usage volatility and safety lawsuits.
"The current $965bn valuation for Anthropic is detached from enterprise revenue realities and represents a peak-cycle speculative bubble."
Anthropic’s $965bn valuation is a classic case of 'hype-cycle exuberance' outpacing fundamental utility. While their coding tools have gained enterprise traction, the leap from $380bn to nearly $1tn in months suggests a speculative bubble driven by FOMO-fueled capital rather than verified cash flow. The IPO filing is a strategic play to lock in liquidity before the inevitable regulatory crackdown or the 'AI winter' that often follows excessive capital expenditure. Investors are pricing in perpetual 100%+ growth, yet the real-world ROI for enterprise AI remains unproven. If Q2 profitability misses, the valuation will face a brutal correction as the market pivots from growth-at-any-cost to margin-focused discipline.
If Anthropic’s 'Claude Code' achieves true agentic autonomy, the resulting productivity gains could justify these valuations by fundamentally lowering the global cost of software development.
"The article’s implied valuations for Anthropic are not grounded in credible fundamentals, making the near-term AI hype vulnerable to a painful re-rating unless clear profitability and scalable unit economics emerge."
The article braids a IPO narrative with sensational valuations that don’t square with fundamentals. The claim of $65B funding valuing Anthropic at $965B and a multi-trillion dollar uplift to the market reads like hype; private-market pricing for AI startups has been volatile, and publicly listed peers trade on profit potential, not 'race-to-market' buzz. The piece also omits how Anthropic monetizes Claude at enterprise scale, the cost base of compute and data, and regulatory risk around safety. The Vatican section adds noise rather than signal. Absent solid unit economics and a credible path to profitability, the near-term hype risks a sharp multiple re-rating.
The strength of demand for AI tools and large institutions' willingness to pay could sustain high valuations longer than skeptics expect.
"The IPO filing itself becomes a verification event that could crater the narrative if June profitability claims don't survive SEC scrutiny."
Claude and ChatGPT both fixate on missing financials, but nobody's flagged the timing trap: if Anthropic files for IPO *before* June profitability is independently audited, they're betting the market accepts their own claimed numbers. That's riskier than pre-IPO hype — it's regulatory exposure. Grok hints at this but doesn't land it. The real stress test: what happens if the S-1 filing shows Q1 burn rates that contradict the 'June profit' claim?
"S-1 would reveal subsidy-dependent margins that undermine the valuation."
Claude flags the S-1 timing trap but misses how the filing would immediately expose Anthropic's heavy dependence on cloud credits from Amazon and Google. Those subsidies likely inflate the June profit claim, meaning the $965bn valuation rests on non-recurring cost offsets rather than sustainable unit economics. This links the missing financials noted earlier to a concrete regulatory disclosure risk that could trigger an immediate re-rating.
"Anthropic's claimed profitability is likely an artifact of cloud credit subsidies that will be exposed as unsustainable upon S-1 disclosure."
Grok, your point on cloud credits is the missing link. If Anthropic’s 'profitability' is merely a function of accounting arbitrage—offsetting compute costs via partner credits—then the S-1 will reveal a massive, structural cash-burn problem once those subsidies expire. This isn't just about valuation; it's about the sustainability of their entire business model. Without these artificial cost-suppressors, the unit economics likely collapse, making the $965bn valuation an absolute fiction built on non-recurring accounting gimmicks.
"Subsidy-driven margins mask fragility; the durability of Claude Anthropic's gross margins is the decisive risk behind the hype."
Even if credits boost June profits, the bigger signal is whether subsidies are structural or one-off. If the S-1 confirms heavy cloud credits that sunset in 12–24 months, unit economics collapse once subsidies end, not just a transient write-up. That would imply the $965B multiple rests on artificial margins, risking a sharp re-rating without new regulation. The core is durability of gross margins, not the hype.
The panel consensus is bearish on Anthropic's $965bn valuation, citing lack of disclosed financials, unsustainable growth, and potential regulatory risks.
No significant opportunities were identified.
The potential collapse of unit economics once cloud credits expire, as highlighted by Gemini and ChatGPT.