AI Is Used by 88% of Companies. But Only a Few Turn It into a Competitive Advantage

Artificial intelligence is no longer a topic reserved for technology enthusiasts or innovation teams. It is becoming a normal part of everyday work. According to The state of AI in 2025, 88% of respondents say their organization regularly uses AI in at least one business function. Generative AI is used by 79% of organizations.

Behind these numbers, however, lies an important detail: most companies are using AI, but have not yet embedded it deeply into how they operate. In other words, AI is present in companies, but it rarely functions at the core of key processes and tends to remain in the testing phase.

Roughly two-thirds of organizations are still in the experimentation or pilot stage. Some are testing their first applications, others are validating specific use cases, but only a minority are actually scaling AI across the organization. According to the research, just 7% of companies have fully deployed and integrated AI across the entire enterprise.

From a consulting perspective, this is the key message: competitive advantage no longer comes from simply “having AI.” The difference starts when a company integrates AI into its core processes in a measurable way.

AI Agents: Strong Interest, Cautious Deployment

One of the most discussed topics in the research is AI agents – systems that can do more than answer a question. They can help with IT service desk support, searching and processing internal knowledge, preparing materials, or assisting with more complex business tasks.

Interest is high, and the data reflects that: 62% of respondents say their organization is at least experimenting with AI agents, and 23% are already scaling them somewhere in the company. Even so, this is not yet widespread practice. In no individual business function does the share of organizations scaling AI agents exceed 10%.

Agents are most commonly used in IT and knowledge management — areas with high volumes of repeated queries, documentation work, internal knowledge retrieval, and service requests. AI agents have strong potential, but it can only be fully realized when they are deployed on the right tasks and have access to relevant data.

AI Is Already Delivering Results, but Often Only in Specific Areas

The research shows that companies are already seeing tangible benefits from AI. Most commonly, these take the form of cost reductions in software engineering, manufacturing, and IT. In these functions, more than half of respondents who use AI report lower costs as a result.

On the revenue side, the greatest benefits appear in marketing and sales, strategy and corporate finance, and product and service development. In marketing and sales specifically, 67% of respondents report revenue growth linked to AI use.

At the same time, the research points to a more sobering reality. Only 39% of respondents attribute any measurable EBIT impact to AI at the enterprise level, and for most of them it does not exceed 5%. AI frequently delivers value at the team or project level, but that impact only partially translates into company-wide financial results.

The Best Companies Do Not Use AI Only to Cut Costs

The research pays particular attention to a group called AI high performers — organizations that attribute more than 5% of their EBIT to AI. They represent roughly 6% of respondents.

Most companies expect AI to deliver efficiency: automation, faster work, lower costs. The most successful organizations, however, do not stop there. They also treat AI as a tool for growth and innovation. Among high performers, 82% cite growth as one of their AI objectives, and 79% cite innovation.

If AI is used only to do the same things more cheaply, much of its potential goes untapped. The real value emerges when companies start asking different questions: How can AI accelerate new product development, open new business lines, or improve management decision-making?

Value Comes from Changing How Work Gets Done, Not from the Tool Itself

One of the strongest findings in the research concerns workflows. Organizations that gain the most value from AI are far more likely to have fundamentally redesigned the way work is done.

Specifically, 55% of AI high performers say they fundamentally redesigned workflows when deploying AI. Among other organizations, the figure is just 20%.

This may be the most important message in the entire research. Buying a license, launching a chatbot, and giving employees access to a tool is not enough. If the underlying way of working does not change, AI remains a cosmetic add-on.

Companies that want to capture more value from AI need to work through the less visible questions: who is responsible for the output, when a human enters the process, how quality is checked, what data AI draws on, how value is measured, and how the new solution makes it into people’s everyday work. Without answering these questions, effective deployment will be very difficult.

Without Leadership from the Top, AI Remains Fragmented

The research also confirms that successful companies deploy AI with stronger support from senior leadership. AI high performers are three times more likely to strongly agree that their senior leaders genuinely own the AI agenda and remain committed to it over time.

This does not mean signing off on a strategy or announcing an AI initiative at a town hall. It means active leadership involvement, regular prioritization decisions, funding, budget reallocation, and leading by example in the use of AI.

AI transformation does not behave like a standard IT project. If leadership leaves it to individual teams, the company ends up with dozens of isolated projects and no real transformation. Treated as a strategic priority, AI can become the backbone of a broader change in how the business operates.

The Impact on Jobs Will Not Be a Simple Story

AI is often discussed in terms of replacing jobs. The research, however, shows a more nuanced picture.

The largest group of respondents, 43%, expects AI to have little or no impact on the total number of employees in their organization over the next year. 32% expect a decrease of 3% or more. By contrast, 13% expect an increase of that magnitude.

At the same time, companies are continuing to hire for AI-related roles. Larger organizations are particularly likely to have hired AI data scientists, data engineers, machine learning engineers, software engineers, and AI product managers.

AI will therefore change not just the number of people, but above all the mix of skills companies need. Some tasks will be automated, others will evolve, and some roles will become more important than before. The question companies will need to answer is not only how many people they need, but what capabilities will be essential in the new way of working.

Risks Grow Alongside AI Use

The more companies use AI, the more they encounter its risks. 51% of respondents from organizations using AI say their company has experienced at least one negative consequence. The most common issue was inaccuracy in AI outputs, reported by 30% of respondents.

Inaccuracy is also the risk companies most actively work to address. Other priority areas include cybersecurity, regulatory compliance, intellectual property protection, and privacy.

Notably, companies reporting the greatest value from AI also report more negative experiences. This does not mean they use AI poorly — more likely, they use it more intensively, across more areas, and often in more sensitive processes. Their exposure to risks is greater, but so is their focus on managing them.

An ambitious AI agenda requires equally ambitious governance. Without clear rules, controls, and accountability, scaling AI becomes very difficult.

Recommendations for Management

AI has become established in business. The question is no longer whether organizations use AI, but whether they use it in a way that changes company-wide performance.

Access to technology is rarely what holds companies back. What separates leaders from the rest is the willingness to invest time and effort into making AI genuinely change the way work gets done. More successful companies treat AI as part of a broader business transformation, not as a standalone tool to cut costs.

The task for management is to move AI from enthusiastic experimentation to managed change: identify priority areas, link them to measurable value, redesign processes, set clear rules, and engage people across the organization.

Want to find out where AI can deliver measurable value in your organization? Reach out to Martin Valášek and explore how to move AI from individual pilots to solutions embedded in your everyday processes.

Source: The research report “The state of AI in 2025: Agents, innovation, and transformation”, QuantumBlack, AI by McKinsey / McKinsey & Company. The online survey was conducted from June 25 to July 29, 2025, among 1,993 respondents from 105 countries