Deloitte China and McKinsey have recently updated their respective reports on enterprise AI adoption in succession. Looking at the two reports together, the data is quite startling.
Deloitte stated: 9% of artificial intelligence projects generate negative returns, and only 4% of enterprises have truly achieved their transformation goals.
McKinsey says: 88% of enterprises have normalized the use of AI in at least one function, but only 6% have become "high-performing enterprises"—that is, those that truly make money from AI, and make a lot of it.
Deloitte looks at China, McKinsey looks at the world; both reports point to the same conclusion: AI is being used with great fanfare, but the real financial returns belong to only a very small number of companies.
Deloitte Index: ROI is generally overestimated
The "2026 Enterprise AI Application Index" jointly released by Deloitte China and the University of Hong Kong surveyed more than 3,200 enterprise IT leaders from 24 countries, and several data points are quite straightforward:
23% Normal use of AI, but business transformation has not yet begun—meaning “used it, but did not get a return”9% AI projects ultimately produced negative returns due to surging hidden costs and business disruption—not only did they not make money, they lost money4% Truly achieved AI transformation goals, with substantive changes to the business model49% AI projects lack a clear definition of success, meaning “done it, but don’t know whether it counts as success”
Deloitte directly points out in its report: enterprises' expectations for return on investment in AI projects are generally higher than actual delivery levels. Some of this falls under “performative AI”—deploying a few chatbots and producing a few AI-generated data dashboards, then publicly claiming to have “completed intelligent transformation,” while contributing less than 1% to EBIT.
Such enterprises account for 23% of Deloitte's sample.
McKinsey’s “6% High-Performance Company” Profile
McKinsey's definition of a "high-performance enterprise" is very strict: EBIT grows by more than 5% due to AI, and AI has already been used to create "significant" value. By this standard, only 6% of companies worldwide meet the criteria.
But what these 6% of companies do is fundamentally different from what the other 94% of companies do:
| Size | High-performance enterprises (6%) | Other enterprises (94%) |
|---|---|---|
| AI Disruptive Transformation Plan | More than three times that of other companies | Most are still in the pilot stage |
| Fundamental restructuring of workflows | 3 times that of other institutions | Overlay AI on top of the original process without changing the process |
| Large-Scale Deployment of Intelligent Agents | More than three times that of other organizations | The vast majority are still in the experimental stage |
| Number of AI use cases | is twice that of other companies | Scattered across several functions, without horizontal scaling |
| Digital budgets are being directed toward artificial intelligence | More than one-third of enterprises allocate more than 20% of their budget to artificial intelligence | Only one-third of enterprises have reached this level |
| AI has been deployed at scale / is being deployed at scale | 74% | 33% |
The logic behind this table is: high-performing enterprises do not simply “use” more AI tools; rather, they restructure business processes, embedding AI at the core of value creation instead of attaching it to the surface of existing workflows.
McKinsey particularly emphasized one detail: high-performing companies more frequently carry out “fundamental restructuring” of their workflows, rather than layering AI capabilities on top of existing processes. In plain terms, this means—you cannot simply slap a layer of AI onto outdated processes and then expect them to become smart.
The Special Status of Chinese Enterprises
McKinsey's report specifically listed data on mainland China, in which several figures differ from the global average:
83% of Chinese enterprises frequently use generative AI in at least one function (global average: 78%); 45% of Chinese enterprises have achieved large-scale or full deployment of AI (global average: 38%); 93% of domestic workplace respondents use AI tools (global average: 58%) — from survey data under the full implementation of "AI+"
Three numbers point in the same direction: Chinese enterprises lead the world in AI penetration, but the proportion of "large-scale deployment" is only 7 percentage points higher than the global average. In other words, many people use it, but not necessarily many have it running smoothly.
This does not contradict Deloitte's index that 'only 4% truly achieve transformation success'—many Chinese companies have already adopted AI tools, but have not yet completed the restructuring of their business processes. In McKinsey's terms, there are still major obstacles between 'routinely using AI in at least one function' and 'achieving large-scale AI deployment.'
Typical Symptoms of “Performative AI”
Reading these two reports together, we can identify several typical symptoms of “performative AI”—see whether your company has also been affected:
- Deploying a chatbot counts as digital transformation: equip customer service, HR, and the IT service desk each with a conversational bot, then issue a press release claiming “full embrace of AI.” According to Deloitte’s data, companies like this account for a considerable proportion.
- AI projects lack a clear definition of success: 49% of companies are unclear about what constitutes “AI project success”—user engagement? cost reduction? or revenue growth? Without a definition, it cannot be measured; without measurement, it cannot be optimized.
- “Pasting AI onto the original process”: McKinsey points out that 94% of non-high-performing companies are doing this. The correct approach is to redesign processes with AI, rather than layering AI tools onto old processes.
- Unaccounted hidden costs: Deloitte points out that 9% of artificial intelligence projects ultimately generate negative returns, mainly because hidden costs such as data cleaning, model fine-tuning, personnel training, and system integration are severely underestimated in the early stages of the project.
Both reports reached a common conclusion: the biggest bottleneck in moving artificial intelligence from “pilots” to “scale” is not technological maturity, but the organization’s execution capacity. The technology is ready, but companies’ internal processes, talent, and incentive mechanisms have not yet caught up.
One-sentence advice for the CIO
If you are planning next year's AI budget, data from Deloitte and McKinsey provide a very direct reference framework:
- First ask: "If this AI project succeeds, how much can EBIT increase?" Then choose the technical route;
- If the answer is “uncertain” or “less than 1%,” then this project is likely just superficial;
- The approach of high-performing enterprises is: select 1-2 core business processes, thoroughly rebuild them with AI, and once successful, expand horizontally;
- Don't launch an AI pilot project in every functional department—94% of companies do exactly that, and it's precisely why most enterprise AI transformations stall.
Ultimately: the endgame of AI transformation is not “we used AI,” but rather “because of AI, our business model has undergone fundamental change.” Deloitte’s 4% and McKinsey’s 6% are the companies that have already reached this point. The remaining more than 90% of companies are still on the way.
Deloitte China × The University of Hong Kong 2026 Enterprise AI Application Index (May 2026) McKinsey 2026 AI Application Status Survey (Chinese version, updated May 2026) China Business Journal (May 23, 2026)
