AI Summit
AI Summit
World Technology Forum — AI Summit
In a single week of June 2026, Anthropic and OpenAI each filed to list at valuations north of most G20 stock markets — while a $700 billion wave of hyperscaler data-centre spending races to keep their models fed. The AI Summit convenes the labs, chipmakers, cloud platforms, sovereign capital and regulators now building — and governing — that economy.
$581.7B
Global corporate AI investment, 2025 — up 130% year on year (Stanford HAI)
46
Companies, platforms and institutions in the directory
12
Countries represented at headquarters level
900M
ChatGPT weekly active users, Feb 2026
Stanford HAI's 2026 AI Index — the field's most cited annual reference — recorded the largest single-year jump in corporate AI investment on record. Every major forecasting house agrees on direction; they disagree sharply on scale, because each is measuring a different slice of the same boom.
Stanford HAI's figure covers private investment, M&A and minority stakes. Gartner's broader "AI spending" measure — which folds in agentic tooling, AI cybersecurity and adjacent data infrastructure — puts 2026 spend at $2.59 trillion; IDC's narrower AI-infrastructure-only figure sits nearer $487–632 billion. None of the three is wrong; they are simply pricing different boundaries around the same industry.
OpenAI, Anthropic, Google DeepMind, Meta and xAI compete on model capability at a compute cost now measured in tens of billions of dollars per training run.
Stargate, hyperscaler capex and sovereign AI vehicles like MGX and HUMAIN are turning compute itself into the industry's scarcest, most contested asset.
Agentic AI — models that take multi-step action rather than answer single prompts — is the fastest-growing enterprise category, led by Agentforce, watsonx and Copilot.
The EU AI Act reaches full application in August 2026, while the US relies on NIST's voluntary framework — a genuine fork in how the world regulates frontier models.
Within a single week, the two most valuable private companies in history both filed to go public. Anthropic filed on June 1, 2026, at a $965 billion valuation following a $65 billion Series H round — having overtaken OpenAI in annualised revenue two months earlier, growing from roughly $1 billion to a $47 billion run-rate in under eighteen months, the fastest revenue scaling in software history. OpenAI followed on June 8 at an $852 billion valuation. Neither had listed as of this Summit's publication, though both are widely expected to be among the largest listings ever completed. A parallel mega-deal reshaped the picture further: Elon Musk merged xAI into SpaceX in an all-stock transaction announced February 2, 2026, valuing the combined entity at $1.25 trillion — and SpaceX itself went public on June 12, 2026, at a reported $1.77 trillion valuation, the largest IPO in history.
Forty-six companies, platforms, financiers and regulators spanning the industry from silicon to sovereign policy. Filter by category, or search by name.
Valuation, revenue and capex figures reflect the most recently reported fiscal results, financing rounds or credible 2025–2026 industry estimates, compiled from public filings, press releases and trade coverage including Stanford HAI's 2026 AI Index Report, CNBC, TechCrunch, Reuters and company investor communications. Private-company valuations move quickly and figures marked as filing or round valuations may not reflect a completed listing. Compiled for Summit convening purposes as of September 2026 and not intended as investment research.
Four threads run through the AI Summit programme, each mapped to the institutions in the directory above.
The handful of labs training the largest models, and the chipmakers and cloud platforms whose capacity gates how fast they can move.
Where model capability meets a P&L — defence, robotics, autonomous vehicles, legal work and banking, including WTFORUM's own OSAI initiative.
Stargate, sovereign AI vehicles and enterprise platform incumbents are converting balance-sheet strength directly into model access.
The EU AI Office, NIST, national AI Security Institutes and ISO/IEC 42001 — the emerging, still-fragmented rulebook for frontier AI.