UNCLASSIFIED // OSINT // ANALYST EYES ONLY
01OVERVIEW
NO FILTER
UTC 2026-05-12 02:22:04Z
SYNC OK
HORIZON2026
DEEP RESEARCH BRIEF

GLOBAL AI // APPLICATIONS, GEOGRAPHIES & MARKET TRAJECTORY

From frontier labs to factory floors. From $725B in hyperscaler capex to $0.10/M tokens. Where AI is being used, who is winning, and where to bet next.

2.5T MARKET$725B CAPEX95% PILOT FAILURE99.7% TOKEN COST DROP15 VERTICALS · 23 COUNTRIES · 200+ SOURCES
GLOBAL AI SPEND 2026
$2.52T
+44% YoY
Gartner
HYPERSCALER CAPEX
$725B
+77% YoY
Tom's Hardware / Statista
ORGS WITH EBIT IMPACT
6%
FLAT
McKinsey
GENAI PILOT FAIL RATE
95%
+2pp YoY
MIT
CHATGPT WAU
850M
+62% YoY
OpenAI
CHEAPEST TIER COST/M
$0.10
−99.7% vs GPT-4
Gemini 3.1 Flash
MCP ENTERPRISE ADOPT.
78%
+47pp YoY
Anthropic / LF
AGENTS IN PRODUCTION
31%
of orgs piloting
S&P Global

HEAD-TO-HEAD // ENT. RUN-RATE ($B)

Anthropic crossed OpenAI in monthly revenue, April 2026
CROSSOVER EVENT
ANTHROPIC $30B+
OPENAI $24B

GLOBAL PULSE // LIVE FEED

STREAMING
  • T-02hMODELDeepSeek V4 Pro / Flash — released under MIT (1.6T params, 1M context)
  • T-04hDEALAnthropic in talks for $50B round at ~$900B post-money
  • T-06hINFRAxAI Memphis Colossus scaling to 1.5 GW by April 2026
  • T-09hAGENTOpenAI Workspace Agents launched, replaces custom GPTs
  • T-12hVERTICALPilot launches fully autonomous AI Accountant
  • T-15hMOVESpaceX completes all-stock acquisition of xAI ($250B)
  • T-18hPOLICYTrump H-1B fee hike to $100K destabilises US talent pipeline
  • T-21hREVENUEAnthropic passes OpenAI in monthly revenue (April 2026)

LAB LEADERBOARD // BY ARR ($B)

#1Anthropic
$30B
#2OpenAI
$24B
#3Google DeepMind
$18B
#4xAI
$5B
#5Mistral
$0.4B
#6Cohere
$0.24B
› DRILL INTO FRONTIER LABS

GEO // TOP 8 BY INVESTMENT INDEX

USA100
United States
CHN72
China
SAU52
Saudi Arabia
ARE48
UAE
IND38
India
KOR34
South Korea
FRA32
France
JPN30
Japan
› OPEN INTERACTIVE WORLD MAP

EXECUTIVE INTEL // THE 8 THINGS THAT MATTER MOST

May 2026 baseline
#01

Anthropic overtook OpenAI on enterprise revenue

$30B+ run-rate by April 2026 (from $1B Jan '25). 40% enterprise spend share vs OpenAI's 27%. Claude Code alone at $2.5B ARR.

#02

The model layer is bifurcating

Frontier reasoning (GPT-5.5, Opus 4.7, Gemini 3 Pro, Grok 4.1) stays expensive & closed. Open-weights bench (DeepSeek V4, Qwen 3.6, Kimi K2.6, Llama 4) matches them at ~10× lower cost.

#03

Coding is the only category with proven unit economics

55% of departmental AI spend. Labs now compete with their own customers (Cursor, Cognition). Legal next: Harvey $190M ARR; Legora fastest-ever to $100M.

#04

Sovereign AI is the defining geopolitical force

UAE, Saudi, India, Korea, Japan, FR, UK, EU, Brazil all building national capacity. Only China attempts full-stack independence — most 'sovereign' efforts run on NVIDIA + US clouds.

#05

Agent layer is real but earlier than the marketing

Agentforce $1.4B ARR. MCP at 78% enterprise adoption. But only 31% of orgs have an agent in prod. 88% of agent pilots stall.

#06

Capex has decoupled from revenue

$725B infra spend vs ~$51B direct AI revenue in 2026. Goldman calls it 'elongation' — buildouts stretch, gap widens, no one blinks.

#07

Pricing is genuinely unsettled

Salesforce runs 3 models in parallel: per-seat, per-conversation ($2), Flex Credits. Intercom Fin's $0.99-per-resolution is the new primitive. IDC: 70% of vendors refactor away from per-seat by 2028.

#08

The largest underbuilt areas aren't where capital is

Vertical AI for regulated industries, blue-collar/field work, non-English markets, agent identity & governance, AI-native services that sell completed work.

SOURCES: GARTNER · IDC · MENLO · MCKINSEY · BCG · BAIN · SACRA · STANFORD HAI · IEA · OECD
DOC ID GAI-26-05 · COPY 001 OF 001
UNCLASSIFIED // OSINT // ANALYST EYES ONLY