# Enterprise AI

> Source: https://aiwiki.ai/wiki/enterprise_ai
> Updated: 2026-06-23
> Categories: AI Tools & Products
> From AI Wiki (https://aiwiki.ai), a free encyclopedia of artificial intelligence. Quote with attribution.

**Enterprise AI** refers to the adoption of [artificial intelligence](/wiki/artificial_intelligence), and since 2023 especially [generative AI](/wiki/generative_ai) built on [large language models](/wiki/large_language_model), inside large organizations: embedded assistants and copilots, autonomous [agentic workflows](/wiki/agentic_ai) driven by [AI agents](/wiki/ai_agent), vertical applications for specific industries, and the governance and security tooling that surrounds them. The field is defined by a persistent tension. Surveyed adoption is close to universal, with 88 percent of organizations reporting AI use in at least one business function by mid-2025 [1], and worldwide AI spending forecast by Gartner to reach $2.59 trillion in 2026 [2]. Yet measured financial returns remain contested, a debate crystallized by MIT Project NANDA's August 2025 finding that about 95 percent of enterprise generative AI pilots produced no measurable profit-and-loss impact, and by the methodological pushback that claim attracted [15][17].

## What does an enterprise AI stack include?

Enterprise AI deployments generally fall into four overlapping layers:

- **Assistants and copilots**: chat-style tools embedded in productivity suites and business software, such as [Microsoft 365 Copilot](/wiki/microsoft_365_copilot), [ChatGPT](/wiki/chatgpt) Enterprise, [Claude](/wiki/claude) for Work, [Gemini for Google Workspace](/wiki/gemini_for_google_workspace), and [Amazon Q](/wiki/amazon_q). A human remains in the loop for every action.
- **Agents**: systems that plan and execute multi-step tasks (resolving a support ticket, processing an invoice, triaging a security alert) with human oversight rather than human operation. Platforms include [Agentforce](/wiki/agentforce), [ServiceNow](/wiki/servicenow) AI Agents, and [Copilot Studio](/wiki/copilot_studio).
- **Vertical applications**: domain products such as [Harvey](/wiki/harvey) in legal, [Abridge](/wiki/abridge) in clinical documentation, and [Sierra](/wiki/sierra_ai) in customer service, which package models with workflow, data, and compliance for one industry.
- **Data, governance, and security platforms**: [retrieval-augmented generation](/wiki/retrieval_augmented_generation) over corporate knowledge (for example [Glean](/wiki/glean)), model gateways, evaluation tooling, and AI governance suites that track usage, risk, and regulatory compliance.

A recurring strategic question is build versus buy. The MIT NANDA study reported that tools purchased from specialized vendors reached successful deployment about 67 percent of the time, roughly twice the success rate of internally built systems [15][16], though many large enterprises pursue a hybrid path: buying horizontal copilots while building agents on differentiating data and workflows.

## How widely is AI adopted in the enterprise?

[McKinsey](/wiki/mckinsey)'s State of AI survey, fielded from June 25 to July 29, 2025 with 1,993 respondents across 105 countries, found 88 percent of organizations using AI in at least one function (up from 78 percent a year earlier), with about two-thirds using it in multiple functions. Only 39 percent attributed any enterprise-level EBIT impact to AI, and most of those put the contribution below 5 percent; roughly two-thirds of adopters remained in experimentation or piloting rather than scaling. McKinsey identified a small cohort of "high performers," about 6 percent of respondents that attributed more than 5 percent of EBIT to AI, capturing disproportionate value [1].

Vendor disclosures point in the same direction of mass adoption. [OpenAI](/wiki/openai) announced on November 5, 2025 that it had passed 1 million business customers, calling itself "the fastest-growing business platform in history," with more than 7 million paid ChatGPT for Work seats, a 40 percent increase in two months, against a base of more than 800 million weekly ChatGPT users [3]. [Anthropic](/wiki/anthropic) reported more than 300,000 business customers by late 2025, with enterprise and API usage described as roughly 80 percent of its revenue [4]. [Microsoft](/wiki/microsoft) said over 90 percent of the Fortune 500 use Microsoft 365 Copilot; paid seats reached about 16 million after its January 2026 earnings call [5] and rose to roughly 20 million by the third quarter of fiscal 2026, the fastest three-month growth since launch [32].

Gartner's May 2026 forecast put total worldwide AI spending at $2.59 trillion for 2026, up 47 percent year over year, with AI infrastructure (chips, servers, and networking) accounting for more than 45 percent of the total, a larger share than software [2].

## What are the main enterprise AI platforms?

| Platform | Vendor | Focus | Reported traction |
|---|---|---|---|
| [Microsoft 365 Copilot](/wiki/microsoft_365_copilot) | Microsoft | Assistant across Office apps; agents via Copilot Studio | Over 90% of Fortune 500; ~20M paid seats by Q3 FY2026 [32]; 230,000+ organizations had used Copilot Studio by April 2025 [6] |
| [Agentforce](/wiki/agentforce) | [Salesforce](/wiki/salesforce) | Autonomous agents on CRM data | $1B+ ARR by Q1 FY2027 (April 2026) [10]; Agentforce and Data 360 near $1.4B ARR, up 114% YoY (Dec 2025) [9] |
| Now Assist / AI Agents | [ServiceNow](/wiki/servicenow) | Generative AI and agents for IT, HR, customer workflows | Crossed $600M annual contract value by end of 2025; $1B targeted during 2026 [11] |
| [AIP](/wiki/palantir_aip) | [Palantir](/wiki/palantir) | Ontology-grounded AI over operational data; "bootcamp" sales model | Q3 2025 revenue $1.18B, up 63% YoY; US commercial up 121% YoY [12] |
| [Glean](/wiki/glean) | Glean | Enterprise search, knowledge graph, agents | $7.2B valuation (June 2025) [13]; ~$200M ARR by late 2025 [14] |

Microsoft 365 Copilot, generally available to enterprises from November 2023 at $30 per user per month [7], anchored the first copilot wave. Salesforce launched Agentforce into general availability on October 29, 2024, initially priced around $2 per agent conversation, marking the major-vendor pivot from copilots to agents [8]. Palantir's Artificial Intelligence Platform (AIP), introduced in 2023, grew through intensive multi-day "bootcamps" in which prospects build working use cases on their own data [12]. Foundation-model providers sell directly into the same market through ChatGPT Enterprise and Claude for Work, while Glean and similar firms compete on cross-application search and agent orchestration.

## How are AI agents used in the enterprise?

The 2025 to 2026 period saw enterprise attention shift from copilots, which assist a person, to agents, which complete work under supervision. McKinsey found 23 percent of organizations already scaling an agentic system somewhere in the enterprise and another 39 percent experimenting [1]. Microsoft reported in February 2026 that 80 percent of the Fortune 500 were running active AI agents [20]. Salesforce said it had closed more than 18,500 Agentforce deals (over 9,500 paid) by December 2025 [9], and Agentforce passed $1 billion in annual recurring revenue in its fiscal first quarter ending April 30, 2026, with revenue up more than 200 percent year over year [10].

Interoperability standards accelerated the shift: the [Model Context Protocol](/wiki/model_context_protocol) (MCP), released by Anthropic in November 2024 and created by engineers David Soria Parra and Justin Spahr-Summers, was adopted by OpenAI in March 2025, Google DeepMind in April 2025, and Microsoft at its May 2025 Build conference, becoming a common way to connect agents to enterprise systems and data. In December 2025 Anthropic donated MCP to the Agentic AI Foundation, a fund under the Linux Foundation co-founded with Block and OpenAI [33].

Forecasters are simultaneously bullish and skeptical. Gartner predicted in August 2025 that 40 percent of enterprise applications would feature task-specific agents by the end of 2026, up from under 5 percent [19], while warning in June 2025 that over 40 percent of agentic AI projects would be canceled by the end of 2027 due to escalating costs, unclear business value, or inadequate risk controls. Gartner also estimated that only about 130 of the thousands of self-described agentic AI vendors were genuine, characterizing the rest as "agent washing": the rebranding of existing chatbots, assistants, and robotic process automation [18].

## Why do most enterprise AI pilots fail?

The sharpest controversy in enterprise AI concerns return on investment. In August 2025, MIT Project NANDA (Networked Agents and Decentralized AI) circulated ["The GenAI Divide: State of AI in Business 2025"](/wiki/mit_genai_divide_report), reporting that despite $30 billion to $40 billion in enterprise generative AI investment, about 95 percent of organizational pilots were producing zero measurable P&L return; only around 5 percent of custom enterprise tools reached production with measurable value [15][16]. The report, based on 52 interviews, 153 leader surveys, and a review of over 300 public deployments between January and June 2025, blamed a "learning gap": generic tools do not adapt to enterprise workflows. As the report put it, "Most GenAI systems do not retain feedback, adapt to context, or improve over time," and it argued the core barrier was "not infrastructure, regulation, or talent" but learning [16]. It also found budgets skewed toward sales and marketing while the strongest returns appeared in back-office automation, and documented a large "shadow AI economy": only about 40 percent of companies had bought official LLM subscriptions, yet workers at more than 90 percent of surveyed companies regularly used personal AI tools [16].

The 95 percent figure spread widely, briefly rattling AI-linked stocks, but drew substantial criticism. Wharton professor Kevin Werbach argued the published material contained no clear support for the headline number and called on NANDA to release its underlying data or retract the report; Futuriom's Scott Raynovich faulted undefined failure criteria, unlabeled charts, restricted access to the document, and the project's vendor sponsorships [17]. Defenders noted the finding measured early P&L impact over a short window, not the technology's potential, and that it matched McKinsey's adoption-to-value gap [1].

Counter-evidence of returns exists at the vendor and case level: ServiceNow and Salesforce monetization figures cited above, [IBM](/wiki/ibm)'s claim of $3.5 billion in productivity gains across more than 70 business areas over two years [30], and [Klarna](/wiki/klarna)'s reported $40 million annual profit improvement from its support assistant [24]. The unresolved question is how far such results generalize beyond well-instrumented early adopters.

## What are the main governance and security risks?

"[Shadow AI](/wiki/shadow_ai)," employee use of unsanctioned tools, is the most pervasive governance problem. A BlackFog survey in November 2025 found 49 percent of employees using AI tools their employer had not approved, with 60 percent saying the security risk was worth meeting deadlines; a third admitted pasting research or datasets into such tools, 27 percent employee data, and 23 percent financial data [21]. IBM's 2025 Cost of a Data Breach report found breaches involving shadow AI cost on average about $670,000 more than other incidents [22]. The canonical early incident came in May 2023, when Samsung banned generative AI tools after engineers pasted internal source code into ChatGPT [23].

Compliance obligations are arriving in stages. The [EU AI Act](/wiki/eu_ai_act) entered into force in August 2024, with general-purpose AI obligations applying from August 2025 and most high-risk requirements originally scheduled for August 2026. In November 2025 the European Commission's Digital Omnibus package proposed simplifying the regime and deferring the high-risk compliance deadline from August 2, 2026 to December 2, 2027, citing delays in designating national authorities and finalizing harmonized standards [34]. Voluntary frameworks, including the [NIST AI Risk Management Framework](/wiki/nist_ai_risk_management_framework) and the ISO/IEC 42001 management-system standard, anchor most corporate AI governance programs, and a tooling market has formed around them (IBM watsonx.governance, Microsoft Purview and Entra controls for agents, and startups such as [Credo AI](/wiki/credo_ai)). As agents gain permissions to act, vendors and analysts increasingly frame agent identity, observability, and least-privilege access as security problems comparable to managing human accounts [20].

## How is enterprise AI affecting jobs?

Klarna is the field's most cited cautionary tale in both directions. In February 2024 it launched an OpenAI-powered assistant that handled 2.3 million conversations in its first month, work it equated to 700 full-time agents, and projected a $40 million profit improvement [24][25]. In May 2025, CEO Sebastian Siemiatkowski told Bloomberg the cost-driven push had produced lower-quality service: "As cost unfortunately seems to have been a too predominant evaluation factor when organizing this, what you end up having is lower quality," he said, adding that Klarna would again recruit human agents for customers who want one while keeping AI as the first line [26].

Salesforce CEO Marc Benioff said in September 2025 that the company had cut its customer support organization from about 9,000 to roughly 5,000 people as Agentforce took over around half of customer interactions, with hundreds redeployed internally [27]; in April 2026 he announced hiring 1,000 new graduates, arguing AI would not eliminate entry-level work [28]. At IBM, CEO Arvind Krishna paused hiring in 2023 for back-office roles he estimated AI could replace (about 7,800 over five years) [29]; by May 2025 he said the AskHR system, which automates 94 percent of routine HR interactions [31], had replaced a couple of hundred HR roles while total employment rose because savings funded engineering and sales hiring [30]. Economists caution that attributing specific layoffs to AI remains difficult, since announcements often coincide with broader cost-cutting; the Klarna, Salesforce, and IBM cases are notable precisely because executives made the attribution themselves.

## References

1. McKinsey & Company. "The State of AI: Global Survey." November 2025. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai
2. Gartner. "Gartner Forecasts Worldwide AI Spending to Grow 47% in 2026." May 19, 2026. https://www.gartner.com/en/newsroom/press-releases/2026-05-19-gartner-forecasts-worldwide-ai-spending-to-grow-47-percent-in-2026
3. OpenAI. "1 million business customers: the fastest-growing business platform in history." November 5, 2025. https://openai.com/index/1-million-businesses-putting-ai-to-work/
4. Sacra. "Anthropic revenue, valuation & funding." https://sacra.com/c/anthropic/
5. AlphaStreet. "Microsoft's 16M Copilot Seats Milestone: Enterprise Adoption or Shelfware Risk?" 2026. https://news.alphastreet.com/microsofts-16m-copilot-seats-milestone-enterprise-adoption-or-shelfware-risk/
6. Microsoft. "Fiscal Year 2025 Third Quarter Earnings Conference Call." April 2025. https://www.microsoft.com/en-us/investor/events/fy-2025/earnings-fy-2025-q3
7. Microsoft. "Furthering our AI ambitions: Announcing Bing Chat Enterprise and Microsoft 365 Copilot pricing." July 18, 2023. https://blogs.microsoft.com/blog/2023/07/18/furthering-our-ai-ambitions-announcing-bing-chat-enterprise-and-microsoft-365-copilot-pricing/
8. Salesforce. "Salesforce's Agentforce Is Here: Trusted, Autonomous AI Agents to Scale Your Workforce." October 29, 2024. https://www.salesforce.com/news/press-releases/2024/10/29/agentforce-general-availability-announcement/
9. Salesforce. "Salesforce Delivers Record Third Quarter Fiscal 2026 Results Driven by Agentforce & Data 360." December 3, 2025. https://www.salesforce.com/news/press-releases/2025/12/03/fy26-q3-earnings/
10. Salesforce Ben. "Agentforce Revenue Surges Past $1B: Here's What You Need to Know." June 2026. https://www.salesforceben.com/salesforce-q1-results-agentforce-hits-1b-arr-as-benioff-takes-aim-at-ai-doubters/
11. FinancialContent / MarketMinute. "AI's 'Spring Awakening': ServiceNow Leads Software Sector Recovery as 'Now Assist' Proves AI Monetization is Real." April 3, 2026. https://markets.financialcontent.com/stocks/article/marketminute-2026-4-3-ais-spring-awakening-servicenow-leads-software-sector-recovery-as-now-assist-proves-ai-monetization-is-real
12. Palantir Technologies. "Q3 2025 Earnings Release" (Form 8-K). November 2025. https://www.sec.gov/Archives/edgar/data/0001321655/000132165525000130/a2025q3ex991earningsrelease.htm
13. Glean. "Glean Raises $150M Series F at $7.2B Valuation." June 10, 2025. https://www.glean.com/press/glean-raises-150m-series-f-at-7-2b-valuation-to-accelerate-enterprise-ai-agent-innovation-globally
14. Futurum Group. "Glean Doubles ARR to $200M. Can Its Knowledge Graph Beat Copilot?" https://futurumgroup.com/insights/glean-doubles-arr-to-200m-can-its-knowledge-graph-beat-copilot/
15. Fortune. "MIT report: 95% of generative AI pilots at companies are failing." August 18, 2025. https://fortune.com/2025/08/18/mit-report-95-percent-generative-ai-pilots-at-companies-failing-cfo/
16. MIT Project NANDA. "The GenAI Divide: State of AI in Business 2025." August 2025. https://mlq.ai/media/quarterly_decks/v0.1_State_of_AI_in_Business_2025_Report.pdf
17. Futuriom. "Why We Don't Believe MIT NANDA's Weird AI Study." August 2025. https://www.futuriom.com/articles/news/why-we-dont-believe-mit-nandas-werid-ai-study/2025/08
18. Gartner. "Gartner Predicts Over 40% of Agentic AI Projects Will Be Canceled by End of 2027." June 25, 2025. https://www.gartner.com/en/newsroom/press-releases/2025-06-25-gartner-predicts-over-40-percent-of-agentic-ai-projects-will-be-canceled-by-end-of-2027
19. Gartner. "Gartner Predicts 40% of Enterprise Apps Will Feature Task-Specific AI Agents by 2026." August 26, 2025. https://www.gartner.com/en/newsroom/press-releases/2025-08-26-gartner-predicts-40-percent-of-enterprise-apps-will-feature-task-specific-ai-agents-by-2026-up-from-less-than-5-percent-in-2025
20. Microsoft Security Blog. "80% of Fortune 500 use active AI Agents: Observability, governance, and security shape the new frontier." February 10, 2026. https://www.microsoft.com/en-us/security/blog/2026/02/10/80-of-fortune-500-use-active-ai-agents-observability-governance-and-security-shape-the-new-frontier/
21. BlackFog. "Shadow AI Threat Grows Inside Enterprises as BlackFog Research Finds 60% of Employees Would Take Risks to Meet Deadlines." November 2025. https://www.blackfog.com/blackfog-research-shadow-ai-threat-grows/
22. IBM. "What Is Shadow AI?" https://www.ibm.com/think/topics/shadow-ai
23. Bloomberg. "Samsung Bans Staff's AI Use After Spotting ChatGPT Data Leak." May 2, 2023. https://www.bloomberg.com/news/articles/2023-05-02/samsung-bans-chatgpt-and-other-generative-ai-use-by-staff-after-leak
24. OpenAI. "Klarna's AI assistant does the work of 700 full-time agents." February 2024. https://openai.com/index/klarna/
25. Klarna. "Klarna AI assistant handles two-thirds of customer service chats in its first month." February 27, 2024. https://www.klarna.com/international/press/klarna-ai-assistant-handles-two-thirds-of-customer-service-chats-in-its-first-month/
26. Bloomberg. "Klarna Turns From AI to Real Person Customer Service." May 8, 2025. https://www.bloomberg.com/news/articles/2025-05-08/klarna-turns-from-ai-to-real-person-customer-service
27. Fortune. "Salesforce CEO Marc Benioff says his company has cut 4,000 customer service jobs as AI steps in." September 2, 2025. https://fortune.com/2025/09/02/salesforce-ceo-billionaire-marc-benioff-ai-agents-jobs-layoffs-customer-service-sales/
28. Fortune. "Salesforce CEO Marc Benioff says AI won't kill entry-level jobs. He's hiring 1,000 new grads to prove it." April 27, 2026. https://fortune.com/2026/04/27/salesforce-ceo-marc-benioff-hiring-1000-new-grads-ai-jobs/
29. Bloomberg. "IBM to Pause Hiring for Back-Office Jobs That AI Could Kill." May 1, 2023. https://www.bloomberg.com/news/articles/2023-05-01/ibm-to-pause-hiring-for-back-office-jobs-that-ai-could-kill
30. Entrepreneur. "IBM Replaced Hundreds of HR Workers With AI, According to Its CEO." May 2025. https://www.entrepreneur.com/business-news/ibm-ceo-ai-replaced-hundreds-of-human-resources-staff/491341
31. IBM. "Transforming HR support with agentic AI" (AskHR case study). https://www.ibm.com/case-studies/ibm-askhr
32. Windows News. "Microsoft 365 Copilot Hits 20M Paid Seats: Enterprise AI Adoption, Governance, ROI." 2026. https://windowsnews.ai/article/microsoft-365-copilot-hits-20m-paid-seats-enterprise-ai-adoption-governance-roi.415952
33. Anthropic. "Donating the Model Context Protocol and establishing the Agentic AI Foundation." December 2025. https://www.anthropic.com/news/donating-the-model-context-protocol-and-establishing-of-the-agentic-ai-foundation
34. Gibson Dunn. "EU AI Act Omnibus Agreement: Postponed High-Risk Deadlines and Other Key Changes." 2026. https://www.gibsondunn.com/eu-ai-act-omnibus-agreement-postponed-high-risk-deadlines-and-other-key-changes/