The State of AI in 2025: Why 88% Use AI But Only 6% See Real Business Impact

McKinsey's latest global survey on the state of AI reveals a striking paradox: while 88% of organizations now regularly use AI in at least one business function, only 6% qualify as "high performers" seeing significant enterprise-wide value. This gap between adoption and impact holds critical lessons for businesses evaluating AI investments.

For companies in Western Canada's energy, construction, agriculture, and mining sectors, these findings illuminate why some AI initiatives deliver transformative results while most remain stuck in perpetual pilot purgatory. More importantly, the research identifies exactly what separates winners from the majority.

The Adoption-Impact Gap: Everyone's Using AI, Few Are Winning

McKinsey surveyed 1,993 participants across 105 nations from June through July 2025, capturing perspectives from organizations of all sizes and industries. The headline finding is simultaneously encouraging and sobering:

Adoption is Widespread:

• 88% of organizations regularly use AI in at least one business function (up from 78% in 2024)
• More than two-thirds use AI in multiple business functions
• 50% use AI in three or more functions
• 62% are at least experimenting with AI agents

But Impact Remains Elusive:

• Only 33% of organizations have begun scaling AI across the enterprise
• Just 39% attribute any level of EBIT impact to AI
• Among those reporting EBIT impact, most say it's less than 5% of total EBIT
• Only 6% qualify as "high performers" (5%+ EBIT impact and significant value from AI)

This pattern creates a critical strategic question: If nearly everyone is using AI but few are capturing significant value, what distinguishes the winners?

What Separates the 6% From Everyone Else

McKinsey's research identifies specific behaviors and practices that high performers employ consistently:

1. Transformative Ambition Over Incremental Gains

High performers are 3x more likely to use AI for transformative change rather than incremental improvements.

While 80% of all respondents cite efficiency as an AI objective, high performers set additional goals:

• Growth and revenue expansion
• Innovation and new product development
• Competitive differentiation
• Market share gains

The research shows organizations pursuing growth and innovation objectives—in addition to efficiency—are more likely to achieve qualitative benefits like improved customer satisfaction, competitive differentiation, and revenue growth.

What This Means for Your Business:

Don't frame AI purely as a cost-cutting tool. While efficiency gains are valuable, organizations limiting AI to cost reduction miss larger opportunities. Ask: "How can AI enable us to serve customers better, develop new offerings, or compete more effectively?" rather than only "How can AI reduce our costs?"

2. Fundamental Workflow Redesign

High performers are nearly 3x more likely to fundamentally redesign workflows rather than simply automating existing processes.

This represents one of the strongest predictors of achieving meaningful business impact among all factors tested. High performers recognize that AI's full value emerges when you rethink how work gets done, not just automate current approaches.

Industry-Specific Examples:

Oil & Gas: Rather than just using AI to speed up existing geological analysis processes, redesign exploration workflows where AI continuously monitors sensor data, identifies anomalies, and triggers targeted human investigation—fundamentally changing geologists' role from data processing to hypothesis testing.

Construction: Instead of using AI to accelerate traditional project scheduling, redesign planning workflows where AI continuously optimizes schedules based on real-time conditions, automatically identifies conflicts, and proposes solutions—transforming project managers from schedule creators to decision-makers evaluating AI-generated options.

Agriculture: Rather than simply automating existing crop monitoring routes, redesign farming operations where AI analyzes continuous satellite and sensor data, identifies issues requiring intervention, and dispatches resources dynamically—shifting farmers from routine monitoring to exception management.

Mining: Instead of just accelerating equipment maintenance schedules, redesign operations where AI continuously monitors asset health, predicts failures, and automatically schedules interventions—moving maintenance teams from reactive/preventive to predictive/prescriptive modes.

3. Scaling Across Multiple Functions

High performers use AI across significantly more business functions than others.

Rather than concentrating AI in a single department, high performers deploy across:

• Marketing and sales (revenue generation)
• Strategy and corporate finance (decision support)
• Product and service development (innovation)
• Software engineering (development acceleration)
• Manufacturing and operations (efficiency)
• IT and service desk (support automation)

Importantly, high performers are also 3x more likely to be scaling AI agents (autonomous multi-step systems) compared to peers.

What This Means for Your Business:

Success doesn't come from having one spectacular AI use case in a single department. It comes from systematic deployment across the organization where each function captures value appropriate to its role. Prioritize building cross-functional AI capability over perfecting a single departmental solution.

4. Senior Leadership Ownership and Commitment

High performers are 3x more likely to report strong senior leadership ownership and commitment to AI initiatives.

This isn't passive support—these leaders actively drive AI adoption, role model AI tool usage, and demonstrate sustained commitment through actions and resource allocation.

The research shows leadership engagement correlates strongly with achieving value from AI. Organizations where leaders actively champion AI and demonstrate personal use see dramatically better results than those where AI remains a middle-management initiative.

Critical Actions for Leadership:

• Actually use AI tools in daily work (Executives asking teams to use tools they don't use themselves undermines adoption)
• Remove barriers to AI implementation (bureaucratic processes, budget constraints, outdated policies)
• Make AI strategy visible through regular communication and strategic planning integration
• Allocate meaningful resources (High performers dedicate 20%+ of digital budgets to AI vs. less for others)

5. Robust Management Practices

High performers consistently implement specific management practices across six dimensions:

Strategy: Clear AI vision integrated into business strategy, not treated as separate technology initiative

Talent: Robust hiring strategies for AI roles (software engineers, data engineers most in demand); continuous upskilling programs for existing workforce

Operating Model: Agile product delivery organizations with well-defined processes; cross-functional teams owning AI outcomes

Technology: Modern infrastructure supporting AI workloads; appropriate tools and platforms for development and deployment

Data: High-quality, accessible data infrastructure (the foundation everything else builds on)

Adoption and Scaling: Defined processes for model validation, human oversight, KPI tracking, and scaling successful pilots

The research shows all these practices correlate positively with capturing value from AI. Organizations implementing more of these practices see better results.

6. Significant Investment

High performers invest more: Over one-third commit 20%+ of digital budgets to AI technologies.

This investment enables them to:

• Scale AI technologies across the business (75% of high performers are scaling vs. 33% of others)
• Hire specialized AI talent (software engineers, data engineers)
• Build robust data infrastructure
• Implement proper governance and monitoring
• Iterate and improve based on learnings

Where Organizations Are Seeing AI Value Today

While enterprise-wide EBIT impact remains rare, the research shows AI delivering measurable benefits in specific use cases:

Cost Benefits Most Common In:

Software engineering - AI-assisted coding, automated testing, code review

Manufacturing - Predictive maintenance, quality control, process optimization

IT - Service desk automation, incident resolution, system monitoring

Revenue Benefits Most Common In:

Marketing and sales - Lead generation, personalization, customer insights

Strategy and corporate finance - Market analysis, forecasting, scenario planning

Product and service development - Innovation acceleration, design optimization

Organizations in these functions should prioritize AI implementation as proven value delivery areas.

The AI Agent Opportunity (And Reality Check)

The research shows significant interest in AI agents—systems that can plan and execute multi-step workflows autonomously:

Current State:

• 62% are at least experimenting with AI agents
• 23% are scaling agents in at least one function
• Most scaling agents in only one or two functions

Where Agents Show Promise:

• IT and service desk management - Automated ticket resolution, system troubleshooting
• Knowledge management - Deep research, information synthesis
• Technology, media, telecommunications - Complex workflow automation
• Healthcare - Administrative workflow automation, research assistance

The Reality: Despite hype, widespread agent deployment remains limited. In any given business function, no more than 10% of organizations are scaling AI agents. This represents emerging capability, not mature technology.

For LootzySoft Clients: We're tracking agent technology developments closely but recommend focus on proven AI applications delivering value today. Deploy agents in controlled pilots where they make sense (service desk, specific research tasks), but don't delay broader AI adoption waiting for agent maturity.

The Workforce Impact: What's Actually Happening

One of the most discussed aspects of AI involves employment effects. The McKinsey research provides nuanced data:

Past Year Reality:

• Plurality of respondents saw little to no workforce change from AI in most functions
• Fewer than 20% reported 3%+ workforce decreases in most functions
• Small percentages added headcount due to AI

Next Year Expectations:

• 30% median expect workforce decreases in functions using AI (up from 17% reporting actual decreases last year)
• 32% predict enterprise-wide workforce reduction of 3%+
• 13% predict enterprise-wide workforce increase of 3%+
• 43% expect no change

Importantly: Most organizations (including larger companies) hired for AI-related roles over the past year. Software engineers and data engineers are most in demand.

The Pattern: AI is changing what people do more than how many people organizations employ. Roles focused on routine data processing and repetitive tasks are declining, while roles requiring AI oversight, strategic thinking, and complex judgment are growing.

For Western Canada Industries:

Expect similar patterns:

• Reduction in routine operational roles (data entry, basic monitoring, repetitive tasks)
• Growth in roles managing AI systems, interpreting outputs, and handling exceptions
• Increased demand for technical talent (data engineers, software developers)
• Premium on workers who can effectively collaborate with AI tools

The strategic response isn't preventing workforce changes—it's managing the transition through reskilling programs and thoughtful workforce planning.

AI Risks: What Organizations Are (and Aren't) Managing

An encouraging finding: Organizations are getting more serious about AI risk mitigation. Respondents now report mitigating an average of four AI-related risks, up from two in 2022.

Most Commonly Mitigated Risks:

• AI inaccuracy
• Personal and individual privacy
• Regulatory compliance
• Organizational reputation

Most Common Negative Consequences Experienced:

• AI inaccuracy (reported by nearly one-third of organizations)
• Intellectual property infringement (especially among high performers with more deployments)
• Regulatory compliance issues

The Gap: Explainability is the second-most-commonly-reported concern but NOT among the most commonly mitigated risks. This suggests many organizations recognize they don't fully understand how their AI systems make decisions but haven't implemented systematic approaches to address this.

Critical Insight: High performers—who deploy twice as many AI use cases—are more likely to report negative consequences. This doesn't mean AI is riskier for them; it means they're deploying AI more extensively and therefore encountering risks others haven't yet faced. Importantly, high performers also mitigate a larger number of risks proactively.

For Your AI Strategy:

Don't treat risk management as optional or something to address later. High performers integrate risk mitigation from the start:

• Define processes for when model outputs need human validation
• Implement monitoring for accuracy, bias, and unexpected behaviors
• Establish governance frameworks before scaling
• Document AI decision-making processes for explainability
• Maintain human oversight, especially for high-stakes decisions

Why Most Organizations Remain Stuck in Pilot Phase

The research reveals two-thirds of organizations haven't begun scaling AI across the enterprise despite widespread experimentation. Several factors explain this:

1. Lack of Data Infrastructure

While not explicitly highlighted in this survey, McKinsey's findings align with other research showing most organizations lack AI-ready data infrastructure. You can't scale AI pilots built on curated datasets when enterprise data is siloed, inconsistent, and poorly governed.

2. Incremental Rather Than Transformative Thinking

Organizations treating AI as "automation of existing processes" capture limited value. Without redesigning workflows to leverage AI's unique capabilities, pilots deliver modest improvements that don't justify enterprise-wide investment.

3. Insufficient Investment

Most organizations aren't investing at levels required for transformation. If you're allocating 5% of digital budget to AI while high performers allocate 20%+, you're not positioning for success.

4. Missing Management Practices

Without agile delivery processes, robust talent strategies, modern technology infrastructure, and systematic adoption approaches, pilots struggle to scale regardless of technical success.

5. Leadership Not Engaged

When AI remains a middle-management or IT department initiative without active senior leadership driving adoption and removing barriers, scaling stalls.

The LootzySoft Perspective: What This Means for Western Canada

McKinsey's research validates what we see working with energy, construction, agriculture, and mining clients:

Start With Data Foundations

High performers implement technology and data infrastructure as a success factor. Organizations with clean, integrated, well-governed data can scale AI pilots; those without this foundation struggle regardless of algorithmic sophistication.

Our recommendation: Before launching multiple AI pilots, invest in data integration, quality controls, and governance frameworks. This foundational work enables successful scaling.

Think Transformation, Not Automation

The 3x difference in workflow redesign between high performers and others represents the single biggest opportunity. Don't ask "How can AI speed up our current processes?" Ask "How should work be done if AI capabilities exist?"

Our approach: We help clients reimagine workflows around AI capabilities rather than simply automating existing processes.

Deploy Across Functions, Not Just One Department

Value comes from systematic deployment across the organization. Marketing, operations, finance, engineering—each function should capture AI value appropriate to its role.

Our recommendation: Develop cross-functional AI capability rather than perfecting single departmental solutions. Build organizational AI literacy and deployment capacity.

Secure Leadership Commitment

Without active senior leadership ownership, AI initiatives stall. Leaders must demonstrate commitment through resource allocation, personal AI tool usage, and consistent strategic communication.

Critical question: Do your senior leaders actively use AI tools in their work? If not, how can you expect organization-wide adoption?

Implement Robust Management Practices

Success requires systematic approaches to strategy, talent, operating model, technology, data, and scaling. Organizations trying to shortcut these dimensions struggle to capture value.

Our role: We help implement the infrastructure, governance, and processes high performers use to scale AI successfully.

Practical Steps: Moving From Pilot to Impact

Based on McKinsey's research, here's how to join the 6%:

1. Assess Your Current State (Honestly)

• Are you in experimentation/pilot phase or genuinely scaling?
• Do you have transformative ambitions or just incremental improvement goals?
• Are workflows fundamentally redesigned or just automated?
• Does leadership actively champion AI or delegate to middle management?
• Is your data infrastructure AI-ready?

2. Set Transformative Objectives

Expand beyond efficiency to include:

• Revenue growth through AI-enabled offerings
• Innovation acceleration in product/service development
• Competitive differentiation through AI capabilities
• Market share gains from superior customer experience

3. Redesign Workflows Around AI

For each AI use case, ask:

• If we could do this work differently knowing AI capabilities exist, how would we?
• What decisions can AI make automatically vs. require human judgment?
• How do we organize work so AI handles what it does well while humans focus on what they do well?

4. Build Cross-Functional Deployment Capability

Don't perfect AI in one function. Build organizational capacity to deploy AI systematically:

• Establish AI centers of excellence providing expertise across functions
• Create standard processes for AI development, validation, and deployment
• Implement governance frameworks enabling responsible scaling
• Develop training programs building AI literacy organization-wide

5. Invest at High-Performer Levels

If you're serious about transformation, allocate resources accordingly. High performers dedicate 20%+ of digital budgets to AI. This funds:

• Robust data infrastructure
• AI talent acquisition and development
• Technology platforms and tools
• Iteration and continuous improvement

6. Implement Risk Management From the Start

Don't wait until scaling to address risks. High performers mitigate risks proactively:

• Define validation processes for model outputs
• Implement monitoring for accuracy and unexpected behaviors
• Establish explainability frameworks
• Maintain human oversight for high-stakes decisions
• Document compliance with regulatory requirements

Conclusion: The Path to the 6%

McKinsey's research delivers a clear message: AI adoption is now table stakes, but capturing significant value requires deliberate strategy and execution. The 6% of organizations seeing transformative impact aren't lucky—they're systematically implementing practices that enable success.

For Western Canada businesses in energy, construction, agriculture, and mining, the opportunity is significant but requires moving beyond experimentation to genuine transformation. The high performers show the way:

• Think big: Set transformative objectives beyond incremental efficiency
• Redesign workflows: Don't just automate existing processes
• Scale systematically: Deploy across multiple functions with robust management practices
• Invest meaningfully: Allocate resources matching your ambitions
• Lead actively: Senior leadership must own and drive AI adoption
• Build foundations: Modern data infrastructure enables everything else

At LootzySoft, we help organizations build the data foundations, implement the management practices, and develop the cross-functional capabilities that high performers use to scale AI successfully. The difference between the 88% using AI and the 6% capturing significant value isn't luck—it's systematic execution of proven practices.

The question isn't whether to use AI—88% already are. The question is whether you'll join the 6% who transform their businesses with it.

Primary Reference: "The State of AI in 2025: Agents, innovation, and transformation"

• Source: McKinsey & Company / QuantumBlack
• Research: Global survey of 1,993 participants across 105 nations
• Survey Period: June 25 - July 29, 2025
• URL: https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai