What should enterprise leaders know about AI today?
On November 12, 2026, Jensen Huang resigned as Nvidia’s chief executive, with CFO Colette Kress named interim CEO and the board promising no interruption to customer commitments. A day earlier United Airlines said it will bar personal AI agents from booking flights, Google priced a new Gemini Decisions API at $0.10 per million input tokens, and on November 12 Deloitte launched Edisyl from an acquired AI team. Leaders should check supplier concentration and decide which AI agents may act for customers.
Executive takeaways
- EvaluateAI Infrastructure: Nvidia lost its founder-CEO overnight while carrying an estimated $400 billion in obligations to other firms; CIOs and CFOs with Nvidia-dependent capacity contracts should review delivery and counterparty terms before the November 17 results.
- EvaluateAI Agents: United is the first US airline to say it will restrict third-party personal AI agents from booking; customer-facing companies need a stated policy on which agents may transact for their customers.
- MonitorAI Economics: Google’s Decisions API scores up to 1,000 questions per request at $0.10 per million input tokens with no output charge, a cheap option for routing, policy checks and grading inside AI workflows.
Nvidia’s CEO resigns, putting AI supplier concentration on the board agenda
AI Infrastructure · NVIDIA, CNBC, November 12, 2026 Priority: Evaluate
What changed?
Nvidia said on November 12, 2026, that Jensen Huang had resigned as president, CEO and director, effective immediately. CFO Colette Kress is interim CEO while remaining CFO, Harvey C. Jones becomes the board’s first chair, and a search covers internal and external candidates (NVIDIA).
The board said it expects operations, customer commitments and the product roadmap to continue without interruption. The release gives no reason; it followed days of reporting on Huang’s personal conduct and an internal investigation, and results remain set for November 17 (SEC).
CNBC reports Nvidia carries about $400 billion in obligations to other firms, including $36 billion in off-balance-sheet backstops for neoclouds as of July. CoreWeave fell about 12% and Nebius about 17% on November 12, while AMD gained (CNBC).
Why does this matter for enterprises?
Much enterprise AI capacity runs through Nvidia directly or through neoclouds whose financing leans on it. A leadership change at the dominant supplier raises execution risk on roadmaps, allocation and those backstops, even if nothing changes this quarter.
This does not mean Nvidia deliveries are at risk today. Analysts quoted by CNBC describe the roadmap as set through 2028; the open question is decision-making over the next several years.
What should leaders do next?
Decision for CIOs, CFOs and procurement leaders: How exposed is our AI roadmap to one chip supplier and the neoclouds it supports?
Within 30 days, map each AI capacity contract to its underlying chip supplier and financing, and identify one workload you could run on an alternative platform.
Sources: NVIDIA — NVIDIA Announces CEO Transition · SEC — Nvidia Form 8-K · CNBC — Nvidia CEO Jensen Huang resigns: live updates
United Airlines bars personal AI agents from booking, keeping the customer relationship in its own channels
AI Agents · Skift, Engadget, November 11, 2026 Priority: Evaluate
For: Retail & Consumer
What changed?
United Airlines said on November 11, 2026, that it will prohibit unauthorized third-party personal AI agents from booking on its platforms and through online travel agencies and aggregators. Skift calls it the first US carrier to say so explicitly; agents can still search and show prices (Skift).
United cited credential sharing, unintended purchases or MileagePlus redemptions, and DOT and FAA rules such as passenger confirmation of bag contents at check-in. Chief executive Scott Kirby told a Chicago event, per The Wall Street Journal, “I don’t want to create another middleman” (Engadget).
Why does this matter for enterprises?
Agentic commerce now has a policy fork: Southwest offers a ChatGPT plug-in, while United, JPMorgan Chase and Amazon restrict outside agents. For airlines, retailers and travel companies, the choice decides who owns the customer data and the transaction.
This does not mean agent traffic is large. Engadget cites Uber reporting 500 agent-placed Uber Eats orders and four rides in a week of 320 million trips and deliveries, and United says agent bookings are minimal.
What should leaders do next?
Decision for CEOs, CDOs and CISOs at customer-facing companies: Which AI agents may act for our customers, and on what terms?
Within 60 days, publish a written agent-access policy covering identity, liability for mistaken purchases and data use.
Sources: Skift — United Airlines moves to block personal AI agents · Engadget — United is restricting AI agents from booking flights
Google prices its Gemini Decisions API at $0.10 per million input tokens, cutting the cost of AI checks
AI Economics · Google, November 11, 2026 Priority: Monitor
What changed?
Google launched the Gemini Decisions API on November 11, 2026, running a Gemini Decision 1 model that scores options instead of writing text. One request can ask up to 1,000 yes/no, multiple-choice or rubric questions about a shared input of up to 256,000 tokens, each with a confidence score (Google).
Pricing is $0.10 per million input tokens with no charge for output, and Google cites 150 to 250 millisecond latency. Google says the model leads its own benchmark suite and is nearly five times cheaper than Gemini 3.5 Flash-Lite for these tasks, without publishing the underlying numbers.
Why does this matter for enterprises?
Agent workflows make many small judgments: which model to call, whether a document passes policy, whether an answer is grounded. A dedicated scorer is faster and cheaper than a full generative model.
This does not mean the scores are reliable for regulated decisions. Google’s benchmarks are self-reported, and the post does not say whether the API is preview or generally available; Google Cloud support is “coming soon.”
What should leaders do next?
Decision for CTOs and heads of AI platforms: Which routing, grading or policy checks in our AI workflows could move to a cheaper scoring model?
Within 60 days, test one high-volume check, such as model routing or content classification, against your current setup and compare accuracy, latency and cost per thousand decisions.
Deloitte turns an acquired AI team into Edisyl, selling business context to the AI tools companies already run
AI Investments & Initiatives · Deloitte, November 12, 2026 Priority: Monitor
What changed?
Deloitte said on November 12, 2026, that it bought the AI business of Flipside Crypto in July and formed Edisyl, a Boston company inside its D71 Ventures unit. Terms were not disclosed; Flipside’s 24-person team joined Edisyl (Deloitte).
Edisyl sells Knowledge Packs: expert-annotated domain knowledge plus a company’s own terms and policies, designed to work across AI systems. It is building first packs with design partners, including unnamed Fortune 500 organizations.
Why does this matter for enterprises?
Large advisory firms are moving from AI consulting hours to owned AI products. Deloitte’s bet is that enterprise AI falls short for lack of business context, not model capability, which matches what many pilot teams report.
This does not mean Edisyl is proven: there are no named customers, results or prices.
What should leaders do next?
Decision for CIOs and CDOs: Who owns the curated business context our AI tools use, and should it sit with a vendor or in-house?
Within 90 days, catalogue the glossaries, policies and definitions your AI assistants rely on, and assign an owner before buying any third-party context product.
Sources: Deloitte — Deloitte announces acquisition of Flipside Crypto’s AI business and launches Edisyl
What should enterprise leaders do next?
| Action | Owner | Timeframe | Success measure |
|---|---|---|---|
| Map AI capacity to its chip supplier and financing | CIO with CFO and procurement | 30 days | every AI capacity contract tagged by underlying supplier and backstop, with one workload validated on an alternative platform |
| Publish an AI agent access policy | CDO with CISO and general counsel | 60 days | a public policy stating which agents may transact for customers and who bears liability for errors |
| Price your AI checks separately from generation | CTO | 60 days | cost per thousand routing or policy decisions measured on your current model and one dedicated scoring model |
The bigger picture
- Control points are shifting: United is guarding its booking channel from outside agents just as Nvidia’s leadership change reminds buyers how much depends on one supplier. Both reward companies that know who sits between them and their customers or capacity.
- AI value is moving into the layers around the model: Google’s scoring API and Deloitte’s Edisyl both sell the checks and context that make models usable, not more model capability.
Executive FAQ
What does Jensen Huang’s resignation mean for Nvidia customers?
Nvidia says operations, customer commitments and the roadmap continue without interruption under interim CEO Colette Kress; the near-term risk is execution and financing support for neoclouds, not current deliveries.
Can AI agents book flights on United Airlines?
No. On November 11, 2026, United said it will prohibit unauthorized third-party personal AI agents from booking, including through travel agencies and aggregators, though agents may still search fares.
How much does the Gemini Decisions API cost?
Google charges $0.10 per million input tokens and nothing for output, with up to 1,000 questions per request.
About the curator & AI disclosure
Curated by Luciano Ayres
Luciano Ayres is Director of Artificial Intelligence at ArcTouch, with more than 20 years of experience in software and technology leadership and specialized studies in deep learning. He focuses on helping organizations connect AI capabilities with practical business outcomes, responsible adoption, and scalable implementation. Connect with him on LinkedIn to discuss enterprise AI opportunities, strategy, investment decisions and risks.
AI-generated briefing. This issue was produced through an automated AI research and writing workflow curated by Luciano Ayres. Sources are linked for independent verification; summaries and analysis may contain errors and do not replace professional due diligence.