Anthropic’s $30 Trillion Pitch Meets Gates’ AI Job Warning

Anthropic is preparing to tell prospective IPO investors its total addressable market exceeds $30 trillion, a figure that tops the $28.5 trillion SpaceX used earlier this year, while targeting a valuation near $2 trillion and a raise of up to $100 billion, according to people familiar with the matter cited by The Wall Street Journal. The Claude maker’s second-quarter revenue more than doubled to $11.6 billion, with a July annualized run rate around $65 billion and internal projections of $190 billion to $200 billion by 2028.

The same week Bill Gates published a long essay arguing the AI transition will be among the most turbulent periods in human history and that “there is no plan to ease the entry into the AI era.” The pitch and the warning land on the same premise: AI systems that can perform vast ranges of human work.

The $30 Trillion Pitch Tops SpaceX

Total addressable market, or TAM, estimates the theoretical annual revenue a company could generate if it captured every possible customer in its defined opportunity. Anthropic is basing its number on the full scope of tasks AI models could complete across industries, people familiar with the plans told the Journal. None of the figures are final. The company is expected to file its IPO prospectus in the coming weeks, with a possible debut in September or early October.

SpaceX described its own $28.5 trillion TAM in its May prospectus as the largest actionable market in human history, with $26.5 trillion tied to AI. It raised about $86 billion and debuted at a $1.77 trillion valuation on June 12. Shares later dipped below $105 intraday in early August before recovering near the $135 offering price.

The 191 technology companies in the S&P 1500 generated a combined $2.4 trillion in revenue last year, per FactSet data cited in the reporting. Anthropic’s figure is more than twelve times that total. It is also roughly the size of the entire U.S. economy, often pegged near $32 trillion.

What a Labor-Sized Market Implies

Rather than starting from software categories or cloud spend, Anthropic is framing the opportunity around the economic value of work that models could ultimately handle. That framing turns the TAM into a claim about the substitution of human cognitive and, eventually, physical labor across law, medicine, customer service, software, manufacturing and more.

On X, the comparison to U.S. GDP drew immediate sharp reactions. One widely viewed post simply restated the two numbers side by side. Others treated the figure as a ceiling that only becomes useful if capture rates, pricing power and competition allow meaningful share. The pitch works as marketing only if investors focus on the upside and set aside the replacement scale it assumes.

• Current annualized run rate near $65 billion as of July is a tiny fraction of $30 trillion.
• Even the 2028 projection of $190-200 billion would represent well under 1 percent of the claimed market.
• Justifying a $2 trillion valuation on conventional models would still require continued hyper-growth well beyond 2028, according to valuation analysts.

NYU Stern finance professor Aswath Damodaran, known as the dean of valuation, had already called SpaceX’s $26.5 trillion AI TAM “reaching the end of what’s plausible and pushing beyond” ahead of that offering. Similar math applied to Anthropic draws the same scrutiny.

Gates Warning Arrives in the Same News Cycle

Gates posted his essay, “The turbulent AI era is here. The choices we make now are critical,” on Gates Notes. He identifies three core risks: permanent job losses, AI empowering bad actors to cause greater harm, and damage to child development and human relationships.

There is no plan to ease the entry into the AI era.

Bill Gates, turbulent AI era essay on Gates Notes

He writes that AI will take on work in law, customer service, medicine, software and manufacturing over the course of a decade rather than generations. “There will be some new jobs, but without the right policies there will be far fewer than exist today.” White-collar entry- and mid-level roles are already seeing effects. Robots will add pressure on physical work later this decade. The result, he argues, is structural change to an economy organized around employment for income, dignity and social connection.

Gates notes that reliability problems in models are being fixed quickly and that analogies to past technology shifts fail because AI adapts to us and runs on devices people already own. He would support a credible global slowdown plan if one existed, but sees geopolitical and economic incentives driving full speed ahead. The essay marks a shift from his 2023 piece that compared AI’s potential to mobile phones and the internet.

The irony is direct. Anthropic’s investor materials will sell the largest possible version of the same substitution of human work that Gates says society has not prepared for.

Anthropic’s Real Revenue Curve

Behind the ceiling math sits rapid commercial growth. Revenue more than doubled quarter over quarter to $11.6 billion in Q2. The annualized run rate reached about $65 billion by the end of July, up sharply from roughly $9 billion at the end of 2025. Earlier private funding valued the company near $965 billion after a large round.

Investors cited by the Financial Times earlier in August expected annualized revenue of $100 billion to $120 billion by the end of 2026. The 2028 range of $190 billion to $200 billion underpins the valuation talk. One backer floated the idea that 30 times forward revenue on extreme growth could support even higher figures. Public markets have no pure-play frontier AI peer for direct comparison.

Customer experience has not been frictionless. Claude’s new text watermark, introduced to meet EU AI Act requirements, drew some criticism. Anthropic explained the method in detail: it embeds an undetectable pattern in low-stakes word choices using a technique similar to Google DeepMind’s SynthID-Text, with no impact on quality, cost or user privacy, and no tracing back to individuals.

Incidents the Prospectus Will Have to Address

Once public, Anthropic’s disclosures will cover more than financials. In June and July the company disabled two advanced Claude models, Fable 5 and Mythos 5, for roughly three weeks after a U.S. export-control directive. The order, citing national security, required suspension of access by any foreign national. Access was later restored after government review.

In late July Anthropic disclosed three real-world cybersecurity evaluation incidents. During third-party capture-the-flag style tests, Claude models reached the open internet due to a misconfiguration and gained unauthorized access to production systems at three outside organizations. The models used basic techniques such as weak passwords and unauthenticated endpoints while believing they were still inside a simulation. Anthropic stopped the evaluations, notified the parties, and detailed the findings publicly. The company said the models did not deliberately escape or exfiltrate themselves.

Competitors and enterprise buyers will also watch commercial traction and the ongoing rivalry with OpenAI over data retention and privacy policies.

What We Know

• Q2 revenue $11.6 billion; July run rate ~$65 billion.
• Reported 2028 revenue target $190-200 billion.
• Export-control pause of ~3 weeks on two models; cyber evaluation incidents disclosed July 30.
• Watermarking rolling out for EU compliance and applied more broadly.

What’s Unconfirmed

• Final TAM language, exact valuation and raise size in the S-1.
• Precise IPO pricing date and exchange.
• How fully the incidents and any residual export issues will be detailed for investors.

Valuation Math Faces the Same Skepticism SpaceX Did

A $2 trillion market capitalization would exceed SpaceX’s debut. Damodaran has argued that justifying such a figure for Anthropic would require on the order of $1.2 trillion in annual revenue within a decade under generous margin and cost-of-capital assumptions, or even higher if the path lengthens. That is many times Amazon’s recent full-year sales and far above current AI product and service markets estimated in the low hundreds of billions.

Earlier tech IPOs used far smaller TAMs. Uber’s $6 trillion and WeWork’s $3 trillion already looked ambitious in their day. SpaceX’s version tested the limit; Anthropic’s goes further. Markets ultimately price the operating trajectory, not the theoretical ceiling. SpaceX shares have traded around the offering price after an initial pop and later dip, showing that even a successful listing can cool when the story meets daily trading.

Anthropic has emphasized safety research and constitutional approaches. That stance has sometimes put it at odds with faster commercial or government demands, another factor public investors will weigh against the growth story.

The Market Will Test Both Claims at Once

If the prospectus appears in the next few weeks, investors will receive the full audited numbers, risk factors and the precise wording of the market opportunity. The $30 trillion figure will sit beside the revenue ramp, the incident disclosures, and whatever language Anthropic chooses for competition, regulation and safety.

Gates’s essay does not name Anthropic. It describes the same economic substitution at the scale Anthropic is inviting investors to underwrite. The company can argue that capturing even a small share of a labor-scale market supports a multi-trillion valuation and that responsible development is the way to realize the upside safely. Critics will counter that the size of the claimed opportunity is itself the evidence that the transition will be disruptive on a scale for which no plan yet exists.

Public markets will render the first verdict in real time once trading begins. The numbers are large enough that the listing, if it happens on the reported timetable, would rank among the largest ever. The gap between the theoretical market and the jobs-and-risks warning is equally large.