A monthly verified-source scan of AI deployments across the 28 industries our matrix tracks. Real outcomes, sourced numbers, no marketing fluff. Generated 2026-05-15.
First in an ongoing monthly series.
This is the first monthly industry scan in what will be an ongoing series. We assessed all 28 industries the matrix tracks for verified AI deployments, real outcomes, and material scoring shifts during April–May 2026. The result is unambiguous.
Zero industries showed no movement. Tech-hardware was the only one with an ADOPTION_PLATEAU flag — specifically on the consumer AI-PC thesis, while the enterprise side simultaneously accelerated.
In Q1 2026, the six largest U.S. banks — JPMorgan, Bank of America, Goldman Sachs, Morgan Stanley, Wells Fargo, and Citigroup — collectively eliminated approximately 15,000 employees while posting combined profits of $47 billion, up 18% year-over-year.
CEOs including Jamie Dimon (JPMorgan) and Charlie Scharf (Wells Fargo) explicitly attributed headcount reductions to AI productivity gains on earnings calls — a qualitative shift from the prior "we're investing in AI" framing.
The proof point that closed the loop: JPMorgan's LLM Suite is now self-funding. 200,000+ employees use it, 150,000 weekly. Software engineers 10% more efficient; operations staff handle 6% more accounts each; fraud cost per unit down 11%. Total annual benefit ~$2 billion — roughly matching the annual AI investment. The "when does AI pay for itself?" question has its first multi-billion-dollar answer.
Combined banking sector headcount is now at its lowest level since 2021. Operations staff at JPMorgan are down 4% while client-facing roles are up 4% — this isn't accidental attrition. It's intentional rebalancing.
Source: The Next Web · April 2026 · The Digital Banker · Q1 2026 · Fortune · Q1 2026 earnings
Beyond banks, four other industries crossed structurally significant thresholds — not incremental progress, but qualitative shifts that change the deployment math.
The fastest large-scale AI adoption curve in any clinical workflow. Microsoft Dragon Copilot and Abridge are dominant. Documented outcomes: 4.15 minutes saved per patient encounter at Cooper University Healthcare (1+ hour per clinician per day); 27% reduction in documentation time at Intermountain Health; 30.7% improvement in documentation-related clinician well-being per JAMA. HCA Healthcare's AI Nurse Handoff is live in 8 hospitals with a 190-hospital system-wide rollout planned (~60,000 daily handoffs). Mayo Clinic launched Platform_Insights to commercialize its AI stack to other health systems globally.
Why it matters: Majority adoption was achieved ~12–18 months ahead of prior scoring projections. The healthcare AI deployment curve needs a material upward revision on the 1–2 year horizon.
Sources: AHA · April 14 2026 · Google Cloud / HCA
Microsoft disclosed 20 million total paid Microsoft 365 Copilot seats in Q3 FY2026 (April 29, 2026) — up from 15 million in January, adding 5 million in a single quarter. Accenture completed a full company-wide rollout to all 743,000 employees on April 27 — the largest single enterprise AI deployment on record. Customers with 50,000+ seat deployments have quadrupled year-over-year.
Simultaneously: ServiceNow launched "Autonomous Workforce" agentic AI completing full IT service desk workflows end-to-end (99% faster than human agents). Salesforce Agentforce reached $800M ARR (+169% YoY) and 29,000 closed deals.
Why it matters: Enterprise software AI is no longer a leading-edge adoption story. It is the mainstream. The question is no longer whether AI deploys at scale — it's what the labor productivity dividend looks like.
Sources: The Next Web · Apr 27 2026 · ServiceNow · May 5 2026 · Fortune · Salesforce
Alphabet's Q1 2026 was structurally unprecedented: Google Cloud grew 63% YoY to $20 billion — the largest cloud quarter ever — with AI-derived revenue up approximately 800% YoY off a smaller 2025 base. Cloud backlog nearly doubled QoQ to $460B+. Gemini Enterprise paid MAUs grew 40% QoQ.
At the same time, Google Network ad revenue declined 4% YoY. This is the first documented case of AI cannibalizing a single company's own legacy product at meaningful scale within one earnings cycle.
Meta separately reported 8 million advertisers using AI creative tools in Q1 (doubled from 4M one year prior); Meta AI assistant crossed 1.2 billion monthly active users.
Why it matters: The comm-services sector needs an internal bifurcation signal distinguishing AI-native revenue streams from legacy ad-dependent streams. Alphabet's internal split is a leading indicator for every other digital ad business.
Sources: Alphabet Q1 2026 SEC filing · Meta Q1 2026
Insilico Medicine's rentosertib (ISM001-055) — a TNIK inhibitor for idiopathic pulmonary fibrosis that was AI-discovered (target identified by AI) and AI-designed (molecule generated by generative AI) — published positive Phase IIa results in Nature Medicine: 98.4 mL improvement in forced vital capacity at 60mg vs. 20.3 mL decline for placebo over 12 weeks. The first AI-generated drug to demonstrate both safety AND efficacy in a controlled human trial. Phase III enrollment in China and Phase IIb in the U.S. are now in preparation.
Recursion Pharmaceuticals simultaneously reported REC-4881 TUPELO Phase 1b/2 positive results (43% polyp burden reduction, deepening to 53% at Week 25) and opened FDA engagement for a registrational pathway. Schrödinger / Nimbus / Takeda's zasocitinib — discovered via AI-aided computational platform — is now in Phase III.
Why it matters: The biotech-AI sector just passed the proof-of-concept threshold. Controlled human efficacy data now exists for a fully AI-generated drug. The probability curve for the entire drug discovery industry shifts.
Sources: Nature Medicine · Insilico · Recursion Q1 2026 earnings
The full per-industry breakdown — flag, headline finding, and direct link to primary source. Sortable mentally. Scannable in 3 minutes.
| Industry | Flag | Headline This Period |
|---|---|---|
| software | DEPLOY_ACCEL | Microsoft Copilot 20M seats; Accenture 743K full rollout; ServiceNow Autonomous Workforce launched. Mass-market threshold crossed. ↗ |
| semiconductors | BIFURCATION | AMD data-center revenue $5.8B (+57%) — surpassed Intel for first time. Broadcom AI $8.4B (+106%); custom ASICs outgrowing GPUs. ↗ |
| pharma | DEPLOY_ACCEL | Merck-Google Cloud $1B partnership; Eli Lilly LillyPod operational (1,016 Blackwell Ultra GPUs); AstraZeneca/Owkin agentic clinical trials. ↗ |
| tech-hardware | ADOPT_PLATEAU BIFUR | Dell COO admits consumer AI-PC "hasn't been what we thought." 50% don't understand the need. Enterprise infra (AI Factory) booming; consumer stalled. ↗ |
| cap-goods | DEPLOY_ACCEL | Honeywell Experion Operations Assistant commercial launch; Caterpillar Cat CS12 autonomous compactor live; Siemens/NVIDIA Industrial AI OS scaling. ↗ |
| comm-services | DEPLOY_ACCEL BIFUR | Google Cloud $20B (+63%); AI revenue +800% YoY; backlog ~$460B; Google Network ad revenue -4% (first self-cannibalization). ↗ |
| div-fin | DEPLOY_ACCEL | Visa Agentic Ready expanded globally; Mastercard/Santander Europe's first live end-to-end AI-agent payment in regulated framework. ↗ |
| banks | DISPL_REAL | 6 largest US banks cut ~15,000 jobs, posted $47B record profits. JPMorgan LLM Suite now self-funding ($2B benefit / $2B cost). ↗ |
| insurance | DISPL_REAL | Optum Real 500M claims YTD (target 2.5B by year-end); Travelers agentic claims voice AI live; ERGO (Munich Re) cutting 1,000 jobs explicitly citing AI. ↗ |
| retailing | DEPLOY_ACCEL BIFUR | Amazon retired Rufus, launched Alexa for Shopping (May 13); Walmart Sparky live + 1,000 corporate cuts via tech-org merger. ↗ |
| media-ent | BIFURCATION | Spotify AI DJ expanded to 75+ markets; NBCU "Live Contextual" AI ads at upfronts; Disney/OpenAI Sora deal collapsed when Sora app shut down. ↗ |
| healthcare | DEPLOY_ACCEL | 62.6% of Epic-using hospitals adopted ambient AI; HCA Nurse Handoff in 8 hospitals (190 planned); Mayo Platform_Insights commercializes AI stack. ↗ |
| autos | BIFURCATION | Waymo 1,400+ sq mi across 11 cities, 500K+ trips/week, $16B raise at $126B valuation. Ford issued 34 recalls in 4 months despite AI factory investment. ↗ |
| energy | DEPLOY_ACCEL | Duke Energy/AWS grid simulation: weeks→15 minutes; NextEra/Google TimesFM grid management; Chevron AI-driven EPS beat. ↗ |
| materials | DISPL_REAL | Dow eliminated 4,500 jobs (~13% workforce) explicitly citing AI; Rio Tinto AutoHaul autonomous rail at full scale; BASF AI sensors -30% batch defects. ↗ |
| telecom | DEPLOY_ACCEL DISPL_REAL | T-Mobile/Ericsson AI-native RAN trials (+10% spectral efficiency); Deutsche Telekom AI Call Assistant launching; Verizon+AT&T shed ~17,700 jobs in 2025. ↗ |
| utilities | DEPLOY_ACCEL | NextEra/Google AI grid management mid-2026 GA; PG&E generative AI at Diablo Canyon (first US nuclear); sector aggregate $1.4T capex through 2030. ↗ |
| transport | DEPLOY_ACCEL BIFUR | UPS RFID network operational across all US small-package operations (Apr 14); Ingram Micro contracts shifted FedEx→UPS citing RFID. ↗ |
| food-staples | DEPLOY_ACCEL | PepsiCo/Siemens digital twins: +20% throughput, -10-15% capex; PepsiCo "agentic AI-first" by end-2026 via Salesforce; Kraft Heinz "Cookbook" AI. ↗ |
| hh-products | DEPLOY_ACCEL | P&G Supply Chain 3.0 global scaling (target $1.5B COGS savings); "unattended shift" manufacturing live in 9 product categories; Kimberly-Clark "Maestro" +$20M savings. ↗ |
| consumer-dur | BIFURCATION | Nike eliminated 1,400 roles (Apr 23) citing automation — 3rd consecutive year of AI-tied cuts; Whirlpool AI appliances launched but Q1 revenue -9.6%. ↗ |
| consumer-svcs | DEPLOY_ACCEL | McDonald's/Capgemini 5-yr renewal across 40K+ restaurants; Starbucks Deep Brew >1B daily transactions; Marriott $1.1B tech investment. ↗ |
| real-estate | DEPLOY_ACCEL | JLL revenue +11% (60% manual lease labor cut, 3x volume with no added headcount); Zillow AI mode launched; CBRE +180bps margin expansion, EPS +81%. ↗ |
| food-retail | DEPLOY_ACCEL | Kroger Gemini Enterprise across 2,730 stores ($400M target); Instacart live inside Claude with 2B+ item catalog; DoorDash AI cuts onboarding 35%+. ↗ |
| biotech-ai | DEPLOY_ACCEL BIFUR | Insilico rentosertib: first AI-generated drug with controlled Phase IIa efficacy (Nature Medicine); Recursion FDA registrational pathway; Schrödinger/Takeda Phase III. ↗ |
| foundry-equipment | DEPLOY_ACCEL | Applied Materials record Q2 (+11%, guides +30% in 2026); KLA +11.5%; ASML guidance raised to €36-40B. AI tools embedded in equipment itself. ↗ |
| chip-design-fabless | DEPLOY_ACCEL BIFUR | NVIDIA data-center $39.1B (+69%); Broadcom AI $8.4B (+106%, line of sight to $100B 2027); Google TPU deal through 2031; AMD MI400 sampling. ↗ |
| household-personal | BIFURCATION | L'Oreal triple-deployment (Beauty Genius 1.1M conversations, K-SCAN +23% salon sales, NVIDIA Alchemi); Estée Lauder +300bps margin via AI supply chain. ↗ |
These are the scoring shifts this scan recommends for the matrix. They get reviewed against existing scores and incorporated where evidence justifies.
Every notable use case in this scan meets three criteria: (1) actually deployed — not announced, not piloted, in production; (2) material to the industry — not a press-release stunt; (3) has a verifiable primary source — Reuters, Bloomberg, FT, SEC filing, company IR, JAMA, Nature, peer-reviewed trade press. Vendor marketing blogs are excluded.
If a section says "no material activity this period," that's a real finding, not padding. The May 2026 scan flagged zero industries with no activity. Future months may.
DEPLOYMENT_ACCEL — Use cases shift the industry's deployment curve faster than current scoring assumes.
DISPLACEMENT_REAL — Concrete labor or revenue displacement evidence (named headcount cuts, lost contracts, public revenue cannibalization).
BIFURCATION — Winners and losers diverging materially inside the same industry.
ADOPTION_PLATEAU — Evidence of slower-than-expected adoption against prior projections.
NO_CHANGE — Nothing material this period.
Monthly. The next scan publishes around the 15th of each month, covering activity from the prior 30-45 days. This series is built so each scan is comparable to the next — same 28 industries, same flag taxonomy, same source standards.
If a finding turns out to be wrong or a source corrects itself, we update the published page and note it. Tell us if you spot a sourcing problem.
This monthly scan is one input to a much larger system. The AI Stocks engine scores 28 industries across 5 horizons (1y, 2y, 3y, 5y, 10y) using 8 analytical dimensions and 167 cross-industry effects — and updates the scores when scans like this one cross structural thresholds (as several did in May 2026).
See the Matrix Scores Free Portfolio ScanAI Is Getting Really Serious — May 2026 reality check: frontier models started training their own successors, the corporate map redrew, the labor impact is asymmetric.
Strange Bedfellows: AI's 18-Month Reshuffle — Ten corporate alliances that should have been impossible. The forcing functions behind every deal.
The Trillion-Dollar Question Wall Street Is Asking Backwards — Why bears and bulls are both right about different questions. How sophisticated capital is pricing it.
The Bagholder Sanity Check — Six cognitive biases picking your AI stocks for you. Self-diagnostic + casino-with-telemetry reframe.