What should enterprise leaders know about AI today?
On November 3, 2026, Menlo Ventures estimated enterprise AI deployment spending at $176.8 billion, with companies spending about 2.5 times what they planned. Goldman Sachs has limited Claude Code and GitHub AI coding tools to engineers after a productivity and risk review. AMD reported data center revenue up 97%, and newspapers sued Brave over a search index sold to AI companies. Leaders should tie AI budgets to measured use and decide who builds software.
Executive takeaways
- EvaluateAI Economics: AI spending is outrunning budgets and concentrating on three labs; CFOs need usage-based forecasts and a second-supplier plan before 2027 budgets close.
- EvaluateEnterprise Adoption: Goldman’s move shows that broad access to AI coding tools is giving way to role-based access with engineering controls; CIOs outside finance face the same choice.
- MonitorAI Infrastructure: AMD’s results and named Helios customers show a credible second source of AI accelerators, which matters for capacity and price negotiations.
Menlo Ventures estimates enterprise AI spending at $176.8 billion, with budgets overrun 2.5 times
AI Economics · Menlo Ventures, November 3, 2026 Priority: Evaluate
What changed?
Menlo estimates enterprise AI deployment spend reached $176.8 billion by September 2026, up 5.4 times from $33 billion in 2025. Surveyed companies on average spent 4.7 times last year’s level against a plan of 1.9 times; the median company spent 2.2 times against 1.8 (Menlo Ventures).
Anthropic, OpenAI and Google take 69% of deployment dollars, up from 50%, and coding tools are a $34 billion market. Menlo credits agents running longer tasks; they generate about 80% of tokens on OpenRouter (GlobeNewswire).
Why does this matter for enterprises?
Agents consume tokens continuously, so AI costs now behave like cloud consumption, not software licenses. Budgets set as a fixed uplift on last year are failing.
This does not mean every company overspent: the survey covers nearly 400 US decision-makers, and Menlo is an investor in Anthropic, which ranks first in its share figures.
| Enterprise function | Business implication |
|---|---|
| Finance | AI needs consumption forecasting and monthly variance review |
| Technology | Agent workloads need per-task cost limits and monitoring |
| Procurement | Supplier concentration calls for a tested second model provider |
What should leaders do next?
Decision for CFOs and CIOs: Do we forecast AI spending by usage, or by a percentage on last year’s budget?
Within 60 days, set token and cost limits per agent workflow, report AI spend monthly against plan, and price one workload on a second provider.
Sources: Menlo Ventures — 2026: The State of Enterprise AI · GlobeNewswire — Enterprise AI spending hits $177B
Goldman Sachs limits Claude Code and GitHub AI tools to engineers, ending open experimentation by traders and bankers
Enterprise Adoption · Bloomberg, The Decoder, November 2, 2026 Priority: Evaluate
For: Financial Services
What changed?
Goldman Sachs updated internal guidance in October to limit Anthropic’s Claude Code and Microsoft’s GitHub coding tools to engineers. Traders and bankers had used them for months to monitor trading flows and produce material, Bloomberg reported on November 2 (Bloomberg).
A Goldman spokesperson cited productivity and risk management, saying the experimentation phase is over. The tools go to staff “best suited to leverage them for the long term”; the firm’s in-house AI tools remain available (The Decoder).
Why does this matter for enterprises?
Software built by non-engineers skips the review, testing and deployment controls a regulated firm relies on. Goldman chose control over broad access.
This does not mean Goldman is retreating from AI. It means AI coding tools are being treated like production engineering, with owners and accountability.
What should leaders do next?
Decision for CIOs, CISOs and chief risk officers: Who in our company may build software with AI agents, and through which controls does that software reach production?
Within 30 days, inventory who uses AI coding tools outside engineering and what they built, and route anything in use through code review.
Sources: Bloomberg — Goldman restricts traders and bankers from using some AI tools · The Decoder — Goldman pulls AI coding tools from traders and bankers
AMD’s data center revenue nearly doubles as Anthropic, Meta, Microsoft, OpenAI and Oracle run on its Helios racks
AI Infrastructure · AMD, November 3, 2026 Priority: Monitor
What changed?
AMD reported record third-quarter revenue of $13.4 billion on November 3, 2026, with data center revenue of $8.5 billion, up 97%. It expects about $17 billion in fourth-quarter revenue (AMD).
AMD says Anthropic, Meta, Microsoft, OpenAI and Oracle Cloud Infrastructure run workloads on its first Helios rack-scale systems.
Why does this matter for enterprises?
A second large accelerator supplier adds capacity and gives cloud providers pricing leverage, which can reach enterprise inference prices.
This does not mean AI compute is getting cheap; the fourth-quarter figure is guidance, and demand still exceeds supply.
What should leaders do next?
Decision for CIOs and CFOs: Are our AI workloads portable enough to use cheaper capacity when it appears?
Within 90 days, ask cloud providers for pricing on AMD-based AI instances and test one inference workload on them.
The New York Times and three newspapers sue Brave over a search index sold to AI companies
AI Regulation & Governance · The Next Web, November 2, 2026 Priority: Evaluate
For: Technology & Telecom
What changed?
The New York Times sued Brave Software in federal court in Manhattan on November 2, 2026, and the Daily News, Chicago Tribune and Orlando Sentinel filed a separate suit. They allege Brave scrapes journalism with an undisclosed crawler, sells access through its Search API and reproduces articles in Ask Brave answers (The Next Web).
The filings name Anthropic, Perplexity, Mistral, xAI and Cohere as users of Brave’s services. The publishers seek damages and an injunction; these are allegations, and no court has ruled.
Why does this matter for enterprises?
Many AI tools ground answers in third-party search indexes. If an index supplier loses, products built on it can change, lose coverage or pass on licensing costs.
This does not create liability for enterprises using AI tools; it raises the value of indemnities and clear data provenance.
What should leaders do next?
Decision for general counsel and CIOs: Do we know which search and data sources our AI vendors use, and are we indemnified if those sources are challenged?
Within 60 days, ask strategic AI vendors to name their web data suppliers and confirm intellectual property indemnity terms.
Sources: The Next Web — New York Times and three other US papers sue Brave over AI scraping
Docusign buys Suplari to connect contract AI with procurement spending data
AI Investments & Initiatives · Docusign, November 3, 2026 Priority: Monitor
What changed?
Docusign announced on November 3, 2026 that it acquired Suplari, a Seattle procurement intelligence company that combines spend, supplier and contract data with AI agents. Terms were not disclosed; Suplari names BT Group, Nordstrom and MediaNews Group as customers (Docusign).
Docusign plans to pair Suplari with its Iris agreement AI to flag off-contract purchases and unmet supplier commitments. These features are not yet delivered.
Why does this matter for enterprises?
Contract terms and actual spending usually sit in separate systems, so savings negotiated are often not realized. Platform vendors are now buying the AI that joins them.
What should leaders do next?
Decision for CFOs and chief procurement officers: How much negotiated contract value do we fail to capture in actual spending?
Within 90 days, measure off-contract spend for one category and use it as the baseline for any procurement AI offer.
Sources: Docusign — Docusign acquires Suplari
What should enterprise leaders do next?
| Action | Owner | Timeframe | Success measure |
|---|---|---|---|
| Move AI budgets to usage-based forecasts | CFO with CIO | 60 days | every agent workflow has a cost limit and monthly spend-versus-plan report |
| Set role-based access for AI coding tools | CIO with CISO | 30 days | 100% of AI-built software in production has passed code review |
| Document AI vendors’ data sources and indemnities | General counsel with CIO | 60 days | data-supplier disclosure and IP indemnity confirmed for every strategic AI vendor |
The bigger picture
- AI is leaving its experiment phase: Menlo’s budget overruns and Goldman’s access limits both show companies replacing open trials with cost and accountability controls.
- Supplier dependence is the new AI risk: concentration on three labs, a second chip source at AMD and the Brave suits all show that what sits under an AI tool shapes its price and continuity.
Executive FAQ
How much are enterprises spending on AI in 2026?
Menlo Ventures estimated on November 3, 2026 that enterprise AI deployment spending reached $176.8 billion by September 2026, up 5.4 times from 2025.
Why did Goldman Sachs restrict AI coding tools?
Goldman cited productivity and risk management when it limited Claude Code and GitHub AI tools to engineers in guidance updated in October, Bloomberg reported on November 2, 2026.
Why are newspapers suing Brave?
The New York Times and three other papers allege Brave scraped their articles and sold access to AI companies through its Search API; Brave has not been found liable.
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. To discuss enterprise AI opportunities, strategy, investment decisions and risks, connect with him on LinkedIn.
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.