AI Investment Statistics 2026
By Axis Intelligence Research
Co-authors: Sarah Mitchell & Simplice Fosso | Last updated: October 28, 2026 | License: CC BY 4.0
Investors put $679 billion into startups worldwide in the first nine months of 2026, the highest total ever recorded for that period, according to Crunchbase. AI companies took 64% of third-quarter funding alone. Two days ago, the first audited financials of a frontier lab became public: Anthropic’s S-1 shows $16.3 billion of revenue in the first half of 2026 against $141.1 billion of capital raised.
Quick Answer: How Much Is Being Invested in AI?
Global corporate AI investment reached $581.7 billion in 2025, up 129.9%, per Stanford HAI’s 2026 AI Index. In 2026, AI captured 82.7% of US venture deal value through September (PitchBook-NVCA). Axis Intelligence Research calculates that Anthropic’s $141.1 billion of raised capital equals 3.1 years of its annualized Q2 2026 revenue.
Key Findings
- Global corporate AI investment reached $581.69 billion in 2025, a 129.9% increase from $253.02 billion in 2024, according to Stanford HAI’s 2026 AI Index (Quid data).
- AI companies took 82.7% of US venture deal value in the first nine months of 2026 — $426.7 billion of $515.8 billion — the highest annual share in the PitchBook-NVCA Venture Monitor’s dataset.
- OpenAI and Anthropic together raised $217 billion in the first half of 2026, equal to 43% of all global startup funding in that period, according to Crunchbase.
- Anthropic’s S-1, filed October 26, 2026, reports $16.29 billion of revenue for the six months ended June 30, 2026, against $1.16 billion a year earlier, and $141.1 billion of capital raised to date.
- According to Axis Intelligence Research, Anthropic’s AI Capital Payback Horizon (ACPH) stands at 3.1 years as of June 30, 2026: $141.1 billion raised ÷ $46.2 billion of annualized Q2 2026 revenue.
How Much Was Invested in AI in 2025?
$581.69 billion, the largest annual total in the Stanford HAI series. That figure is corporate AI investment: mergers and acquisitions, minority stakes, private funding rounds and public offerings involving AI companies, measured by Quid for the Stanford HAI 2026 AI Index economy chapter. It is not enterprise AI spending and not hyperscaler capital expenditure.
| Component (2025) | Value | Change vs 2024 | Source |
|---|---|---|---|
| Total corporate AI investment | $581.69B | +129.9% | Stanford HAI / Quid |
| Private investment | $344.66B | +127.5% | Stanford HAI / Quid |
| Mergers & acquisitions | $214.44B | +132.6% | Stanford HAI / Quid |
| Generative AI private investment | $170.87B | more than +200% | Stanford HAI / Quid |
| AI infrastructure, models, research & governance (focus area) | $143.22B | largest focus area | Stanford HAI / Quid |
| Private funding events above $1B | 28 (15 in 2024) | +13 events | Stanford HAI / Quid |
| Average private AI funding event | $66.5M | +46% | Stanford HAI / Quid |
The decade view matters more than the single year. Stanford’s series starts at $14.57 billion in 2013, so 2025 is roughly 40 times the 2013 level. The path was not smooth: corporate AI investment peaked at $360.73 billion in 2021, fell to $201.00 billion in 2023, and only passed the 2021 peak in 2025.
Sarah Mitchell’s read: the 2025 number is less about breadth than about where the cheques landed. The infrastructure-and-models bucket alone took $143.2 billion, about 42% of private AI investment by Axis’s division of Stanford’s figures. Capital went to the layer that buys compute, not the layer that sells applications.
Which Countries Invest the Most in AI?
The United States, by a margin no other country approaches. US private AI investment was $285.88 billion in 2025, 23 times China’s $12.41 billion and about 14 times Europe’s combined $20.92 billion (Stanford HAI).
| Country / region | Private AI investment, 2025 | Newly funded AI companies, 2025 | Source |
|---|---|---|---|
| United States | $285.88B | 1,953 | Stanford HAI / Quid |
| Europe (aggregate) | $20.92B | — | Stanford HAI / Quid |
| China | $12.41B | 161 | Stanford HAI / Quid |
| United Kingdom | $5.90B | 172 | Stanford HAI / Quid |
| France | $4.36B | 84 | Stanford HAI / Quid |
| Canada | $4.28B | 79 | Stanford HAI / Quid |
| India | $4.09B | 108 | Stanford HAI / Quid |
| Germany | $3.89B | 92 | Stanford HAI / Quid |
| Israel | $3.58B | 64 | Stanford HAI / Quid |
| Saudi Arabia | $2.03B | — | Stanford HAI / Quid |
China’s private figure understates its total. Stanford estimates that government guidance funds directed about $184 billion into AI firms between 2000 and 2023, money that never appears in private-market tallies. Read the 23-to-1 ratio as a gap in private capital, not in national commitment. Our China AI statistics page breaks down the state channel.
Within the US, California alone accounted for $218 billion of 2025 private AI investment, more than 75% of the national total (Stanford HAI).
How Much AI Venture Funding Has Been Raised in 2026?
More than in any full year before it, and the year has a quarter left. The three main measurement firms disagree on the exact total, so this page publishes all of them rather than picking one.
| Measure | Q1 2026 | Q2 2026 | Q3 2026 | 2026 YTD | Source |
|---|---|---|---|---|---|
| Global venture funding | $305B (as of July 1 data) | $212B (revised) | $159B | $679B | Crunchbase |
| Global VC investment | — | $227.4B | $170.9B | $728.3B | KPMG Venture Pulse |
| US VC deal value | — | — | $98.4B | $515.8B | PitchBook-NVCA |
| AI share of US deal value | 88.4% | 84.1% | 65.9% | 82.7% | PitchBook-NVCA |
| AI share of global funding | — | more than 70% | 64% ($102B) | — | Crunchbase |
Sources: Crunchbase Q3 2026 global report (data as of Oct. 2, 2026); KPMG Venture Pulse Q3’26 (Oct. 14, 2026); PitchBook-NVCA Venture Monitor Q3 2026 (as of Sept. 30, 2026). Crunchbase’s Q2 figure was $205 billion in its July report and $212 billion by October; early tallies drift upward as late-reported deals arrive.
Why do Crunchbase and KPMG disagree on Q3 2026?
Because they count different things. KPMG put Q3 2026 at $170.9 billion across 8,341 deals; Crunchbase put it at $159 billion across close to 6,000 startups. Axis Intelligence Research calculates the gap at $11.9 billion, or 7.5% of the Crunchbase figure, and $49.3 billion (7.3%) on the year-to-date totals. Neither firm is wrong; the definitions of a venture round, the treatment of corporate rounds, and reporting lags differ.
One detail worth knowing before citing a US number: KPMG’s US figure for Q3 ($98.4 billion across 3,626 deals) matches the PitchBook-NVCA US deal value to the decimal, which suggests a shared underlying dataset. Crunchbase counted $91 billion for US-based startups in the same quarter. If you need two independent US readings, use Crunchbase and PitchBook-NVCA, not KPMG and PitchBook-NVCA.
Is AI’s share of venture funding falling?
Within 2026, yes — and that is mostly arithmetic, not retreat. PitchBook-NVCA’s AI share of US deal value fell from 88.4% in Q1 to 65.9% in Q3 because the frontier-lab mega-rounds that dominated the first half did not repeat. Crunchbase says no single company raised in the tens of billions in Q3.
Breadth went the other way. A record 27 companies raised rounds of $1 billion or more in Q3 2026, up from 16 in Q2 and 14 in Q1, and those 27 took about a third of global venture capital (Crunchbase). The largest were Databricks and Safe Superintelligence at $5 billion each.
Sarah Mitchell’s read: when two companies stop raising, the whole market’s AI share drops by twenty points. A falling share in Q3 is a measure of how much of H1 was two cheques, not a sign that investors cooled on the category.
How does 2026 compare with 2025 for AI venture funding?
US AI deal value through September 2026 was $426.7 billion, about double the $214.1 billion recorded for all of 2025 (PitchBook-NVCA). Total US deal value of $515.8 billion is roughly 44% above 2021’s full-year record of $357.9 billion. Globally, Crunchbase’s nine-month total of $679 billion already exceeds the $440 billion it recorded for all of 2025.
Who Is Getting the Money? Concentration in AI Funding
Two companies. Crunchbase reports that OpenAI and Anthropic together raised $217 billion in the first half of 2026, 43% of all global startup funding. Anthropic alone took close to a third of Q2 2026 global venture funding with its $65 billion Series H.
| Company | Round (2026) | Amount | Post-money valuation | Source |
|---|---|---|---|---|
| OpenAI | Closed March 31, 2026 | $122B committed | $852B | OpenAI announcement (company-reported) |
| Anthropic | Series G, February 2026 | $30B | $380B | Crunchbase News |
| Anthropic | Series H, May 28, 2026 | $65B | $965B | Anthropic announcement (company-reported) |
| Databricks | Q3 2026 | $5B | — | Crunchbase |
| Safe Superintelligence | Q3 2026 | $5B | — | Crunchbase |
| Mistral AI | Q3 2026 | $3.5B | — | KPMG Venture Pulse |
| Moonshot AI | Q3 2026 | $3.5B | — | KPMG Venture Pulse |
| Crusoe | Q3 2026 | $3.9B | — | KPMG Venture Pulse |
Concentration holds even without the two frontier labs. PitchBook-NVCA reports that deals of $100 million or more were about 5% of US deal count in Q3 2026 but 75% of deal value, and J.P. Morgan, writing in the same Venture Monitor, puts a record 68% of capital deployed through nine months into the top 10 deals.
The valuation premium is visible at every stage. PitchBook-NVCA reports a median AI valuation step-up of 2.2x in 2026 against 1.5x for non-AI companies.
What Did the Money Buy? Anthropic’s S-1 and the AI Capital Payback Horizon
Until October 26, 2026, every answer to this question relied on run-rate figures leaked to the press. Anthropic’s S-1 changed that: it is the first registration statement from a frontier model lab with financial statements, filed for a Nasdaq Global Select listing under the ticker ANT.
| Anthropic, PBC (S-1, filed Oct. 26, 2026) | Value |
|---|---|
| Revenue, 2023 / 2024 / 2025 | $41M / $386M / $4,585M |
| Revenue, H1 2025 → H1 2026 | $1,163M → $16,286M |
| Revenue, Q2 2026 | $11,555M (+1,368% year over year) |
| Compute and infrastructure expense, H1 2026 | $8,945M |
| Operating loss, H1 2026 | $(693)M |
| Adjusted operating income, H1 2026 (non-GAAP) | $1,618M (10% margin) |
| Net loss, H1 2026 | $(80,748)M, of which $(72,388)M is a change in fair value of convertible notes, derivatives and SAFEs |
| Capital raised to date (June 30, 2026) | $141.1B |
| Cash, equivalents and short-term investments (June 30, 2026) | $131.65B |
| Net revenue retention (Sept. 30, 2026) | ~900% |
The net loss headline will travel furthest and mislead most. About 90% of the $80.7 billion is a non-cash revaluation of convertible instruments, driven by Anthropic’s own rising valuation. The operating loss for the half was $693 million.
What is the AI Capital Payback Horizon (ACPH)?
The AI Capital Payback Horizon (ACPH) is an Axis Intelligence Research metric: capital a company has raised to date, divided by its annualized latest-quarter revenue. It answers one question in years: how long would the company need, at its current revenue pace, to bring in as much revenue as investors have put in? It measures revenue, not profit, and it does not predict returns.
Formula: ACPH = cumulative capital raised ÷ (latest quarterly revenue × 4)
| Company | Capital raised | Revenue basis | Annualized revenue | ACPH | Source tier |
|---|---|---|---|---|---|
| Anthropic (as of June 30, 2026) | $141.1B | Q2 2026 revenue $11.555B × 4 | $46.22B | 3.1 years | Filed (S-1) |
| Anthropic, TTM basis | $141.1B | Q3 2025–Q2 2026 revenue | $19.71B | 7.2 years | Filed (S-1) |
| OpenAI (estimate, March 31, 2026 revenue) | $181.9B (PitchBook estimate) | $2B per month, company-reported | $24B | 7.6 years | Estimate |
| OpenAI (estimate, end-September revenue) | $181.9B (PitchBook estimate) | ~$50B annualized, told to investors (CNBC) | ~$50B | ~3.6 years | Estimate |
Axis Intelligence Research calculation, October 28, 2026. Anthropic inputs from its S-1; OpenAI capital from the PitchBook-NVCA Venture Monitor Q3 2026 (“total VC raised to date”); OpenAI revenue from its March 31, 2026 announcement and from CNBC’s October 8, 2026 report on an OpenAI investor presentation. The OpenAI rows mix measurement levels and are labeled estimates.
Which revenue figure goes into the denominator matters. CNBC reported that OpenAI told investors it reached roughly $50 billion of annualized revenue at the end of September 2026, below a $68 billion figure circulated weeks earlier that, according to a person familiar with the matter, included gross revenue from OpenAI’s partners. The same report says Anthropic told investors its run rate hit $65 billion at the end of July; on that unaudited basis, Anthropic’s ACPH would read about 2.2 years. The 3.1-year baseline uses only filed revenue.
The baseline reading is 3.1 years for Anthropic, the only frontier lab with filed figures. Two things move it. Revenue growth shortens it fast: on the TTM basis, which weights older, smaller quarters, the same company reads 7.2 years. New capital lengthens it: Anthropic’s IPO proceeds, once priced, will add to the numerator.
A reconciliation worth stating: PitchBook estimates Anthropic’s total VC raised at $124.3 billion, while the S-1 reports $141.1 billion of capital raised. The $16.8 billion difference is likely definitional — the S-1 figure covers all capital, including convertible notes such as those held by Amazon, which a venture-round database may not count as VC. When citing, the filed number wins.
Is AI revenue keeping up with AI investment?
At the frontier lab level, revenue is now outrunning compute cost. Axis Intelligence Research divided Anthropic’s revenue by its compute and infrastructure expense for each period in the S-1:
| Period | Revenue | Compute & infrastructure | Revenue per $1 of compute |
|---|---|---|---|
| 2023 | $41M | $433M | $0.09 |
| 2024 | $386M | $2,522M | $0.15 |
| 2025 | $4,585M | $7,326M | $0.63 |
| H1 2026 | $16,286M | $8,945M | $1.82 |
Axis Intelligence Research calculation from Anthropic’s S-1.
The 2026 crossover — more than a dollar of revenue for every dollar of compute — is the first filed evidence that a frontier lab’s unit economics can turn. It covers one company, during a half-year in which revenue grew 14-fold. A single filing is a data point, not an industry.
Stanford’s chapter adds the other side of the ledger for 2025: OpenAI’s reported annual compute spend was $16.3 billion and Anthropic’s $6.8 billion (Epoch AI data).
How Are Hyperscalers Financing AI: Equity, Debt or Cash Flow?
Increasingly with debt. Venture rounds are only one channel; Big Tech funds most AI infrastructure through capital expenditure, and that spending now exceeds cash generation at some companies.
Meta’s third-quarter 2026 results show capital expenditures of $34.91 billion against $32.79 billion of operating cash flow, leaving free cash flow of negative $2.12 billion. Meta’s long-term debt rose to $80.92 billion on September 30, 2026 from $58.74 billion at the end of 2025, after $24.91 billion of debt issuance in the first nine months. Meta narrowed its 2026 capex outlook to $135–145 billion.
| Financing signal (2026) | Value | Source |
|---|---|---|
| Hyperscaler investment-grade bond issuance, year to date | ~$225B, more than the previous four years combined | J.P. Morgan, in PitchBook-NVCA Venture Monitor Q3 2026 |
| Debt raised by US VC-backed companies, through Q3 | $81.7B (vs $69.7B in all of 2025) | PitchBook-NVCA |
| Meta long-term debt, Sept. 30, 2026 | $80.92B (Dec. 31, 2025: $58.74B) | Meta Q3 2026 release |
| Corporate venture capital, global, year to date | $532.8B, above the prior annual record | KPMG Venture Pulse Q3’26 |
Company-by-company capex guidance, filings and the cash-flow ratio we track every quarter sit on the AI Capex Tracker; data center sites and megawatts are on the AI Data Center Tracker.
Sarah Mitchell’s read: the venture numbers on this page are the equity layer of a much larger stack. When a hyperscaler’s capex exceeds its operating cash flow, the gap is filled by bonds, and the bond market — not venture capital — becomes the price-setter for the next round of AI compute.
Where Is AI Investment Going in 2027?
Worldwide AI spending is forecast to reach $3.64 trillion in 2027, up 36.2% from $2.67 trillion in 2026, according to Gartner’s September 16, 2026 forecast. No measurement firm has published a 2027 venture forecast that Axis Intelligence Research could verify, so the rest of this section reports dated, sourced signals rather than an invented number.
| Gartner AI spending by market | 2025 | 2026 | 2027 (forecast) |
|---|---|---|---|
| AI infrastructure | $981.9B | $1,484.4B | $1,977.7B |
| AI services | $434.0B | $576.5B | $745.7B |
| AI software | $288.2B | $461.6B | $656.4B |
| AI cybersecurity | $25.9B | $51.3B | $86.0B |
| AI agents and assistants | $16.5B | $29.2B | $65.5B |
| Generative AI models | $13.0B | $28.3B | $51.6B |
| Total AI spending | $1,786.7B | $2,670.5B | $3,637.3B |
Source: Gartner, September 2026 (Table 1, selected markets; total includes all markets). Growth rates and shares are Axis Intelligence Research calculations.
Three readings from the table. Infrastructure is 55.6% of 2026 AI spending and stays above half (54.4%) in 2027 — the buildout continues, but its share is beginning to slip. AI agents and assistants are the fastest-growing market Gartner lists for 2027, at 124%. And spending dwarfs investment: Stanford’s 2025 corporate AI investment total ($581.7 billion) equals about 22% of Gartner’s 2026 spending forecast — capital raised by AI companies is a fraction of the money customers and hyperscalers spend on AI products and infrastructure.
The exit window opens through Anthropic first. Anthropic’s S-1 is public; PitchBook-NVCA notes that OpenAI has ruled out a 2026 listing. CNBC reports that OpenAI confidentially filed its prospectus in June and that executives have signaled a 2027 debut, while OpenAI holds early talks on a new round of around $30 billion. Whichever lab lists first sets the public-market price for frontier AI revenue, and every private mark below it will be re-read against that price in 2027.
Liquidity is the bottleneck. The US has 992 active unicorns worth a combined $5.7 trillion, and PitchBook-NVCA estimates the IPO backlog at more than two full years of listings. US VC funds raised $108.5 billion through Q3 2026, with megafunds of $500 million or more taking 78.1% of it.
Hyperscaler spending keeps rising. Meta told investors it expects “to continue investing aggressively” in 2027, without giving a figure. Gartner forecasts AI infrastructure spending of $1.98 trillion in 2027, up 33.2%; company-level capex guidance is on our AI Capex Tracker.
Rates are the swing factor. PitchBook-NVCA warns that rising rates could add pricing pressure through 2027, and the J.P. Morgan commentary in the same report ties the AI capex cycle directly to bond-market conditions.
The near-term outlook is still positive. KPMG expects Q4 2026 venture investment to “remain buoyant,” led by AI infrastructure, physical AI and robotics, while flagging a possible pre-midterm slowdown in the US.
Methodology
Collection. Every figure on this page comes from a document fetched on October 28, 2026: Stanford HAI’s 2026 AI Index economy chapter (Quid and Epoch AI data); Crunchbase News quarterly reports for H1 and Q3 2026; KPMG Venture Pulse press releases for Q2 and Q3 2026; the PitchBook-NVCA Venture Monitor Q3 2026 PDF; Anthropic’s S-1 (SEC EDGAR, filed Oct. 26, 2026) and Series H announcement; Meta’s Q3 2026 earnings exhibit on SEC EDGAR; and Gartner’s September 16, 2026 AI spending forecast. Two OpenAI revenue figures are company-reported (March 31, 2026) or press-reported (CNBC, Oct. 8, 2026, on an investor presentation) and are labeled as such.
Definitions. “Corporate AI investment” (Stanford) includes M&A, minority stakes, private investment and public offerings. “Venture funding” (Crunchbase, KPMG, PitchBook-NVCA) counts equity rounds into startups and differs by firm on corporate rounds, stage classification and data cutoffs. We never add figures from different firms.
Axis calculations. The AI Capital Payback Horizon (ACPH) = capital raised ÷ (latest quarterly revenue × 4). Revenue per compute dollar = revenue ÷ compute and infrastructure expense, both from the S-1. Measurement-spread figures are simple differences between published totals. All inputs and results are rows in the CSV with axis_calculated = yes.
What this page does not measure. Enterprise AI spending (see the AI statistics page), hyperscaler capex company by company (see the AI Capex Tracker), and sovereign AI programs, for which we found no measurement-firm dataset we could verify this cycle.
About This Dataset
ai-investment-statistics.csv holds every figure on this page as one row with provenance columns (source_org, source_document, source_url, retrieved_date, is_primary, axis_calculated, method_note). Values are raw numbers with units in a separate column; dates are ISO 8601. Released under CC BY 4.0. The page is revised when Q4 2026 venture tallies publish, when Anthropic’s IPO prices, or when Stanford HAI releases its 2027 AI Index.
How to Cite This Page
APA: Axis Intelligence Research, & Mitchell, S. (2026, October 28). AI investment statistics 2026. Axis Intelligence. https://axis-intelligence.com/ai-investment-statistics/
MLA: Axis Intelligence Research, and Sarah Mitchell. “AI Investment Statistics 2026.” Axis Intelligence, 28 Oct. 2026, axis-intelligence.com/ai-investment-statistics/.
Chicago: Axis Intelligence Research and Sarah Mitchell. “AI Investment Statistics 2026.” Axis Intelligence, October 28, 2026. https://axis-intelligence.com/ai-investment-statistics/.
AI Investment FAQ
Is AI investment in a bubble?
The filed data cuts both ways. Anthropic’s S-1 shows revenue up 1,368% year over year in Q2 2026 and $1.82 of revenue per dollar of compute in H1 2026, but its last private valuation ($965 billion) is about 21 times annualized Q2 revenue, by Axis calculation. Meanwhile PitchBook-NVCA counts 992 US unicorns worth $5.7 trillion waiting on an IPO window that has not broadly reopened.
How much did OpenAI and Anthropic raise in 2026?
OpenAI closed $122 billion in committed capital at an $852 billion post-money valuation on March 31, 2026, per its announcement. Anthropic raised $30 billion in its February Series G and $65 billion in its May 28 Series H at $965 billion post-money. Together the two took 43% of global startup funding in H1 2026, per Crunchbase.
What share of venture capital goes to AI?
It depends on the tracker and the quarter. PitchBook-NVCA puts AI at 82.7% of US deal value through September 2026, a record. Crunchbase puts AI at 64% of global venture funding in Q3 2026, down from more than 70% in Q2 because the largest frontier-lab rounds were not repeated.
Why is China’s AI investment so much lower than the United States’?
Private markets are only part of China’s AI financing. Stanford HAI records $12.41 billion of Chinese private AI investment in 2025 against $285.88 billion in the US, but estimates that government guidance funds sent about $184 billion into Chinese AI firms between 2000 and 2023, outside private-market data.
Does venture funding include Big Tech AI capex?
No. Venture totals from Crunchbase, KPMG and PitchBook-NVCA count equity rounds into startups. Hyperscaler capital expenditure — Meta reported $34.91 billion in Q3 2026 alone — is a separate, balance-sheet channel tracked on Axis Intelligence Research’s AI Capex Tracker.
Is Anthropic profitable?
Not on a GAAP basis. Anthropic’s S-1 reports a $693 million operating loss and an $80.7 billion net loss for H1 2026, most of it a non-cash revaluation of convertible instruments. On its own non-GAAP measure, adjusted operating income was $1.6 billion, a 10% margin.
How much will be spent on AI in 2027?
Gartner forecasts worldwide AI spending of $3.64 trillion in 2027, up 36.2% from $2.67 trillion in 2026, with AI infrastructure alone at $1.98 trillion. That is spending on AI products and infrastructure, not capital invested in AI companies; no measurement firm has published a verifiable 2027 venture forecast.
What is the best single number for “AI investment”?
There isn’t one. Use Stanford HAI’s $581.7 billion for 2025 corporate AI investment (deals and funding), a measurement firm’s venture total for startup funding in 2026, and capex filings for infrastructure spending. Mixing them double-counts and mislabels.
Explore More Axis Intelligence Research
- AI Capex Tracker — every hyperscaler capex disclosure, dated and primary-sourced
- AI IPO Tracker — filings and pricings for AI listings, including Anthropic and OpenAI
- AI Statistics 2026 — AI spending, adoption and ROI
- Generative AI Statistics — generative AI adoption and enterprise spend
- China AI Statistics — China’s state and private AI capital
- AI Data Center Tracker — sites, megawatts and leases behind the investment
- AI ROI Calculator — model the return on your own AI spend
Correction, October 28, 2026
This page was rebuilt after an audit against our Editorial Directive V9. The June 22, 2026 version contained errors, now corrected:
- It described Anthropic’s $30 billion round as a “Series H” at a $965 billion valuation. Anthropic’s $30 billion round was its February Series G at $380 billion; its Series H was $65 billion at $965 billion (May 28, 2026).
- It stated OpenAI’s annual revenue run rate as $3.4 billion and Anthropic’s as $2 billion in early 2026. OpenAI reported $2 billion of revenue per month on March 31, 2026; Anthropic’s S-1 reports $4.73 billion of revenue in Q1 2026 alone.
- Its 2014–2024 trend table understated corporate AI investment in every year. Stanford HAI’s series shows, for example, $19.04 billion in 2014 (not ~$4 billion) and $360.73 billion in 2021 (not $93.5 billion).
- It stated Cerebras opened at $350 versus a $185 IPO price as a “68% premium”; that arithmetic gives 89%, and the figure has been removed as unverified.
- It reported an “AI Investment Concentration Index” reading of 91.4 as “the highest since 2022.” No historical series existed; the index has been retired and replaced by the AI Capital Payback Horizon, published here as a baseline.
- Unverified figures (including several Q1 2026 deal counts, sovereign-AI totals and sector funding totals) have been removed.
Report an error: [email protected] · Corrections policy