“AI is now used almost everywhere, but very few organizations are capturing real value from it.” — Alex Singla, Senior Partner at McKinsey
Have you ever wondered how far most companies really are with AI? According to the latest global survey by McKinsey & Company, conducted between June 25 and July 29, 2025, with 1,993 participants from 105 nations, 88% of organisations report using AI in at least one business function. Yet only about one-third say they are scaling AI across the enterprise.
While many teams are testing new tools, running small pilots, and experimenting with AI agents , the gap between interest and real impact remains wide. The survey shows that 62% of companies are exploring or using agents, but only 39% report any noticeable improvement in profit. Many organisations face the same issues. Limited data quality, unclear goals, slow adoption inside teams, and rushed tech decisions make it hard to turn pilots into strong business value. This reflects a common problem. AI is growing fast, but real success depends on the systems, habits, and clarity behind it.
In this blog, you will learn what the current state of AI looks like, what leading companies are doing differently, and how your organisation can move from quick trials to results that matter.
Key Takeaways Choose an AI development company based on business fit, domain expertise, and alignment with your KPIs. Look for end-to-end technical capability, strong integration skills, and solid MLOps practices . Prioritise partners with clear data governance , strong security standards, and experience in compliance. Favour vendors with structured delivery, transparent communication, and tangible proof of execution. Ensure the team has a balanced skill mix across data, ML, engineering, and product. Review pricing clarity, IP ownership, and long-term support terms before committing. Test partners through a small, KPI-driven proof of concept before scaling.
Transform Your Business with AI-Powered Solutions! Partner with Kanerika for Expert AI implementation Services
Book a Meeting
AI Adoption Continues to Expand Across Industries Artificial intelligence has now moved into the mainstream of global business. According to McKinsey’s State of AI report, companies across sectors are integrating AI into customer operations, marketing, product development , and analytics. Even so, enterprise transformation is still uneven. Most organizations are using AI, but few have reached full-scale deployment across the business. This gap shows that 2025 is becoming a decisive year where companies must evolve from experimentation into operationalized, organization-wide AI adoption.
Mainstream Adoption Shifting From Pilots to Broader Experimentation AI adoption is now nearly universal. McKinsey reports that 88% of organizations use AI in at least one function, reflecting how deeply the technology has entered everyday operations. This includes everything from process automation to generative content support and workflow acceleration.
However, adoption does not yet translate into maturity. Most companies remain in a testing phase, running small pilots or isolated use cases instead of scaling AI across functions. The challenges typically relate to fragmented data, limited governance, and a lack of workforce readiness to support broader rollout.
What this means for organizations AI is widely adopted but not deeply integrated. Scaling remains slow due to infrastructure and operating model limitations. Most deployments exist within departments, not across entire enterprises.
AI is present everywhere, but its impact is still developing.
Leading Agentic AI Startups & Platforms in 2025: What Sets Them Apart Agentic AI companies reshaping enterprise automation with autonomous, multi-agent systems.
Learn More
The Rise of AI Agents Across Enterprise Workflows One of the most transformative shifts highlighted in the report is the rapid rise of AI agents . These systems can perform multi-step tasks autonomously, acting as intelligent co-workers rather than simple response tools.
McKinsey’s findings show strong early movement in this direction. 39% of organizations are experimenting with agents, while 23% have begun scaling them in production environments. This makes AI agents one of the fastest-growing trends in enterprise AI.
“Getting to transform companies requires that you deploy digital and data analytics solutions, but it is just as important to build change management and capabilities so those solutions scale.” — Alex Singla, Senior Partner, QuantumBlack
Where AI agents are gaining traction IT operations: automated troubleshooting, system monitoring, service desk support Knowledge management: documentation, summarization, research, information retrieval Customer operations: intelligent routing and early-stage autonomous support workflows
Industries leading adoption Technology Media and telecom Healthcare
AI agents are shifting the narrative from AI as a support tool to AI as a task owner. This is driving faster productivity gains and reshaping how organizations design workflows.
Ambitious Organizations Are Gaining the Most Value From AI While most companies are still navigating the learning curve, a smaller group of AI high performers is pulling ahead significantly. These organizations have clearer strategies, disciplined governance, stronger investments, and a more transformative approach to AI-enabled work.
High performers are organizations achieving at least a 5% improvement in EBIT attributable to AI. They also report substantial qualitative value, such as improved decision-making, faster development cycles, or higher customer satisfaction. These companies treat AI as a core capability rather than a tactical add-on.
Unlike average adopters, high performers pursue broader outcomes. They use AI to expand revenue, launch new offerings, and differentiate themselves in the market. While efficiency matters, their primary goal is long-term growth and innovation.
They rely on AI to
2. Workflow Redesign Plays a Central Role High performers stand out in how they integrate AI into operations. Instead of adding AI to old processes, they rebuild workflows with AI at the center. This approach allows them to scale faster and unlock deeper operational value.
They focus on
Designing AI-native workflows Removing unnecessary manual steps Integrating data , automation, and decision systems into a unified structure
Source: Mckinsey 3. Leadership and Governance Support Successful Scaling These organizations also benefit from stronger leadership commitment. Their executives drive AI strategy directly, enabling cross-functional alignment and faster decision-making.
They put in place
This structure gives them the stability needed to adopt AI at scale.
High performers invest significantly more in AI. Many allocate over 20% of their digital budgets to AI platforms, data infrastructure , and talent. This level of sustained investment directly correlates with their ability to scale and generate business impact.
Their investment priority areas include
McKinsey’s data shows a clear relationship: greater investment leads to faster scaling and more substantial returns.
Partner with Kanerika for Expert AI implementation Services
Book a Meeting
Expectations Vary on AI’s Effect on Workforce Size As AI adoption continues to expand, organisations remain divided on how it will reshape their workforce. McKinsey’s latest survey shows that expectations are still fluid, with leaders unsure whether automation will replace jobs or simply redistribute work. While some anticipate reductions driven by efficiency, others believe AI will strengthen existing roles and create new technical and operational capabilities. Despite these mixed views, hiring demand for AI talent has increased significantly over the past year, especially in engineering and model-building roles.
1. Mixed Predictions for Headcount Companies report varying expectations for future workforce size.
32% expect workforce reductions as automation increases 43% anticipate no change in total headcount 13% expect workforce growth of at least 3% Uncertainty remains high as only a minority has fully scaled AI across the enterprise
“Organizations now need to be more adaptable. They must move fast but not break things, be intentional on measuring the right risk, and align against the right values to ensure ROI.” — Lareina Yee, McKinsey
2. Function-specific Workforce Trends The expected impact of AI differs widely across business functions.
Operational and support functions are more likely to expect reductions Product, analytics, software engineering, and IT functions anticipate stable or rising headcount Automation-focused industries show stronger expectations of efficiency gains Roles involving creativity, stakeholder communication, and decision-making remain less affected
3. Hiring for AI Skills Continues to Accelerate Even with some predicted reductions, organisations continue investing heavily in AI capabilities.
Growing hiring demand for software engineers, ML engineers, data engineers , and AI product managers Many organisations report increased budgets for AI skill development and internal upskilling Hiring for roles like prompt engineers and AI governance specialists is rising Over half of companies plan to expand their AI or data-focused teams in the coming year
Efforts to Mitigate AI Risks Are Becoming More Common As AI becomes embedded in everyday business workflows, risk mitigation has evolved into a strategic priority. Leaders increasingly recognise that responsible AI is essential for trust, compliance, and long-term value creation. Although organisations have improved their approach to privacy, security, and governance, gaps remain in areas such as model transparency and explainability, which are crucial for regulated industries and high-stakes use cases.
1. Growing Awareness and Mitigation Efforts Awareness of AI risks continues to increase across industries.
More companies are implementing privacy, security, and regulatory safeguards Adoption of governance frameworks and responsible AI guidelines is rising year over year Investment in risk monitoring tools and model audit systems continues to grow
2. Most Common Risks Experienced Despite progress, negative AI outcomes remain widespread.
Around 51% of organisations have experienced at least one adverse consequence Inaccuracy remains the most commonly reported issue across teams Bias, incorrect recommendations, and compliance gaps continue to appear in early-stage deployments Leaders note the need for more consistent validation and human oversight
Source: Mckinsey 3. Gaps in Risk Response Some critical risks remain under-addressed.
Explainability and transparency receive less attention than privacy or security Many models still operate as black boxes, making auditing and debugging difficult Fast deployment timelines often overshadow responsible oversight Limited visibility into model behaviour slows compliance readiness
Organisations with the strongest AI outcomes adopt a more structured and proactive approach.
High performers face more complex risks, including IP exposure and regulatory scrutiny They invest heavily in governance, validation, and human-in-the-loop review Strong leadership involvement accelerates responsible scaling Safeguards are integrated directly into processes rather than added later
Organisations are rapidly embracing AI, but most remain stuck between experimentation and true enterprise transformation. McKinsey’s insights show that while adoption is high, scaling is limited by operational, technical, and organisational barriers. The path forward requires a shift from isolated pilots to end-to-end redesign, stronger governance, and a commitment to building long-term capabilities. High performers demonstrate that AI success is not about tools alone, but about ambition, structure, and readiness for agent-driven automation.
1. The Scaling Challenge Many organisations remain in pilot mode despite rising AI budgets and broader adoption.
A large share of companies still lack fully operationalised AI use cases Legacy systems, fragmented data, and unclear ownership slow enterprise-wide scaling Teams often deploy isolated solutions without connecting them to core workflows Scaling requires rethinking processes, not simply adding more models
High performers follow a structured, disciplined approach that accelerates value creation.
They redesign workflows instead of layering AI onto outdated processes Senior leadership actively sponsors, funds, and monitors AI initiatives Rigorous model validation and continuous monitoring are standard practice They invest early in data foundations, scalable infrastructure, and cross-functional teams Their strategy focuses on growth and innovation, not just cost efficiency
“There’s a real learning curve. The sooner you get started on the learning curve, the quicker you’ll reach higher levels.” — Michael Chui, Senior Fellow, QuantumBlack
AI agents are emerging as a powerful lever for deep automation.
Agents handle multi-step tasks independently, reducing manual intervention Early adoption is rising, with organisations experimenting and beginning to scale They enable more autonomous process execution, especially in IT and knowledge-heavy functions Agentic systems will become key to scaling AI beyond individual use cases
4. Building Capabilities for the Next Phase For true transformation, organisations must strengthen their capabilities across multiple areas.
The Path to AI at Scale: Insights and How Kanerika Enables the Journey McKinsey’s 2025 findings show that AI adoption is widespread, yet actual enterprise-scale impact remains uneven. While nearly 90% of companies now use AI in some form, most remain in pilot mode and only a minority report meaningful EBIT gains. Organizations that outperform treat AI as a transformation lever rather than a set of isolated tools. They redesign processes, strengthen data foundations, and invest in the talent and governance required to scale responsibly. The message is clear: value comes from bold, integrated execution, not experimentation alone.
At Kanerika, we help enterprises move beyond pilots by combining advanced AI solutions with practical automation. Our AI agents, such as Alan for legal document summarization , Susan for PII redaction, and Mike for proofing and data validation, are designed to solve real business challenges. These agents reduce manual effort, improve compliance, and accelerate workflows, making AI adoption tangible and impactful.
We complement this with secure, scalable systems built on robust compliance frameworks, including ISO 27701, ISO 27001, SOC II, GDPR, and CMMI Level 3 . Our expertise spans data integration , analytics, AI/ML, and cloud management, supported by partnerships with Microsoft, AWS, and Informatica. As businesses prepare for advanced AI capabilities like agentic workflows and autonomous systems , Kanerika provides the technology, governance, and collaboration needed to turn ambition into measurable enterprise impact.
Partner with Kanerika for Expert AI implementation Services
Book a Meeting
FAQs 1. What does The State of AI in 2025 highlight? The report shows that most firms now use AI in at least one part of their work. A smaller group uses it across many teams, and these firms tend to get stronger results. The study also notes that many leaders are testing AI agents to handle tasks and support staff. A key theme is that firms that plan well and adjust their ways of working see more value than those that only run small tests.
2. Why is it hard for firms to scale AI? Many firms struggle because they try to fit AI into old systems. True scale needs changes in workflow, better data setups, and clear goals. Some teams also lack the right talent or steady support from top leaders. When these pieces are missing, AI remains stuck in small trials instead of helping across the whole company.
3. How are AI agents used in 2025? AI agents are used to handle tasks such as support tickets, research, planning steps, and simple creative work. Most firms are still in early stages, but interest is rising fast. The strongest results show up when teams reshape their workflow so the agent is part of daily tasks and not just an extra tool on the side.
4. Will AI change jobs in 2025? Some teams say they expect slight reductions in staff while others expect small increases. Many firms report no major change. The biggest shift is that workers spend less time on routine steps and more time checking results, planning ideas, or working with customers. The focus is not on removing jobs but on moving people to higher value tasks.
5. What sets top performing AI firms apart? These firms treat AI as a core part of their plan rather than a side project. They invest more, support teams with training, and reshape workflows so that people and AI work together. They also set clear goals around growth and innovation instead of only trying to cut costs. This mindset helps them roll out more use cases and see stronger results.
