A year ago, most gen AI use cases existed as proofs of concept or early-stage ideas not yet fully implemented in any organization. But the technology is maturing rapidly, and organizations are now deploying agents and gen AI solutions at significant scale: 28 percent of top-performing companies plan to increase their technology budgets by more than 10 percent in 2026 specifically to scale agentic AI systems that autonomously plan and act across workflows.
At McKinsey Technology’s annual CIO Conference, we convened CIOs from all major industries to discuss how this AI revolution is shaping a new era technologically and from a business perspective. Four main themes emerged.
Speed has become a more significant differentiator than size
In the past, organizations have often harnessed economies of scale to become more competitive, the understanding being that the larger the organization, the easier it is to increase efficiency. But with the massive proliferation of agents and the possibility to automate vast amounts of work, scale can be achieved far more readily with automation than through organizational size alone. Large organizations may struggle to move quickly and innovate, but the age of agents requires quick decision-making and clear guardrails to build faster innovation cycles. Smaller organizations may thrive under these conditions. McKinsey research shows that top-performing companies are adopting product and platform operating models where decisions happen within days instead of months.
It is no longer sufficient to pursue a technology agenda in isolation
Traditionally, technology functions handled digitization. Business and technology teams worked through dedicated interfaces, structured around product owners and business owners, to translate operational needs into technical requirements, while the rest of the organization continued to run as before. But today, automation is extending beyond the technology function to include every team on the business side, with agents working alongside employees to automate tasks and streamline workflows. As a result, 46 percent of top performers now cocreate strategy across business and tech teams iteratively throughout the year—a figure that has doubled since 2023.
McKinsey Global Institute research indicates the future of work will be a partnership among people, agents, and robots, all powered by AI. Today’s technologies could theoretically automate more than half of current work hours, making the question of how organizations prepare their people as consequential as the technology decisions themselves.
This paradigm shift requires a fundamental restructuring of both business and technology teams. It also requires reskilling at scale. About half of top performers are already investing in reskilling their own workforces to deliver the next wave of change—setting up operating models that integrate human and agentic work and ensuring teams across the enterprise are equipped to work with agents. An environment where human beings and agents can collaborate and thrive requires incorporating the strengths of human beings while focusing on the opportunity ahead and mitigating risks.
Autonomous capabilities are now deeply embedded in business processes
AI solutions and robots have historically automated parts of existing processes—such as chatbots resolving routine customer queries before passing complex cases to a human agent—with manual workarounds as an option in case technology failed or produced incorrect outputs. But agents are now able to manage numerous steps and interact with each other.
In customer service, for example, agents can classify tickets, identify root causes, propose resolution paths, and autonomously resolve frequent low-complexity issues end to end. These operating model shifts are already happening, with 23 percent of organizations reporting scaling AI agents in at least one business function and an additional 39 percent actively experimenting.
These new operating models no longer allow for manual fallbacks. Therefore, organizations must harden their technology solutions—reducing vulnerabilities and building resilience (for example, by removing unnecessary software), regularly installing security updates, and introducing redundancy so that operations can continue if individual components, such as cloud providers, fail.
Partnering within the ecosystem
Scaling agentic AI extends beyond the boundaries of any single organization. As agents increasingly operate across shared platforms, cloud infrastructure, and third-party systems, the interfaces between organizations become as consequential as the internal workflows they connect. Technology leaders can no longer treat vendor and partner relationships as transactional. Instead, codeveloping solutions end to end with cloud providers, technology vendors, and implementation partners—as well as aligning on architecture, governance, and risk from the outset—will be essential to capturing the full value of the agentic era. Organizations that get this right will find that these partnerships become a source of competitive advantage, not just a procurement necessity.
In this new era of automation, people across functions could have the opportunity to do more meaningful and higher-value work. Realizing this prospect and capturing the full benefits of AI transformations requires establishing smaller, more agile teams with a sustained focus on upskilling employees to work effectively with large-scale agentic systems.
Across all discussions at the conference, one idea was constant: The next wave of technological advancement cannot be achieved through technological excellence alone. Connecting people across disciplines will be critical, and close collaboration among business and technology leaders, domain experts and data scientists, and hardware and software capabilities will ultimately determine whether organizations can convert the potential of agentic AI into trillions of dollars in sustainable annual impact. We are already looking forward to continuing these conversations at next year’s conference.
Anna Wiesinger is a partner in McKinsey’s Düsseldorf office, Henning Soller is a partner in the Frankfurt office, Sven Blumberg is a senior partner in the Istanbul office, and Martina Gschwendtner is a consultant in the Munich office, where Thao Dürschlag is an associate partner.
The authors wish to thank Emily Hoppe, Hannah Hoehnke, Hannah Mayer, Kimberly Beals, Marie Heller, and Sophie Westphal for their contributions to this blog post.