After years of experimentation, agentic AI has proven itself in sandboxes. In 2026, the focus shifts from capability to accountability: How do you govern an autonomous workforce? Who supervises the supervisors? And what happens when unsanctioned “shadow agents” start making decisions nobody approved?
Industry leaders weigh in on what’s coming as autonomous systems move from proof-of-concept to production.
1. Agentic AI moves from pilots to performance
Darin Patterson, VP of Market Strategy, Make
“If 2025 was the year of the AI agent, then 2026 will be the year businesses finally turn AI potential into tangible value. Organizations will move beyond experimentation to reliable AI-powered automation, measuring AI by its impact on efficiency and outcomes.”
Eduardo Crespo, VP EMEA, PagerDuty
“2026 will be the year for deploying AI agents at scale. Fully led AI companies will emerge where AI doesn’t simply assist but drives strategy, innovation and customer experience. Developers are shifting from coders to architects of intelligence, guiding agents and curating what machines create.”
Andy Whitehurst, CTO, Sopra Steria UK
“In 2026 we’ll see agentic AI pilot projects evolve into mainstream adoption across UK enterprises. Increasing levels of trust will see more complex processes being handled with speed and accuracy. This won’t be about replacing humans entirely. Agentic AI will form part of a true hybrid workforce, freeing people to focus on the most complex, empathetic interactions.”
2. Software development leads, consumer-facing use lags
Rob Mason, CTO, Applause
“AI’s biggest 2026 growth area will remain in software development, where technical users are more tolerant of iteration and errors. Developers are managing agentic workforces, transforming them from coders into managers and coaches. Customer-facing applications are progressing more slowly due to reputational and regulatory risks.”
Tony Gentilcore, Cofounder, Glean
“Natural language has replaced syntax as the interface of creation. Developers will stop typing functions and start describing intent. Agents will build, test and validate systems on their own. Humans will design outcomes, not loops.”
Peter Schneider, Principal Product Manager, Qt Group
“The winners won’t be those replacing developers with AI, but those using focused automation strategically. Smart managers will ensure automation supports skill development rather than replacing it.”
3. Multiagent systems and orchestration become foundational
Chris Dyke, Sales Director UK & Ireland, Allied Telesis
“By 2026, early adopters will implement multiagent systems to automate routine tasks and optimize decision-making. These systems will enable hybrid models where AI handles predictable operations and humans focus on complex situations.”
Steven Webb, UK CTO, Capgemini
“Specialized multiagent systems are far better suited to automating end-to-end workflows than single ‘copilot’ use cases. New frameworks are emerging to help organizations orchestrate and coordinate agents safely, but governance and observability remain major barriers.”
Daniel Meyer, CTO, Camunda
“The next chapter of agentic AI is less about how smart any single agent is and more about how well organizations coordinate multiagent interactions, while maintaining the highest levels of governance and accountability.”
Jakob Freund, CEO and Cofounder, Camunda
“Enterprise agentic automation will close the gap between vision and reality by blending dynamic AI with deterministic guardrails and human-in-the-loop checkpoints to deliver trusted autonomy at scale.”
4. Governance, trust and the rise of “shadow agents”
Manoj Nair, Chief Innovation Officer, Snyk
“In 2026, a surge in shadow agentic AI will create one of the largest blind spots in enterprise security. Unsanctioned agents with broad access will act as unmonitored digital insiders, forcing companies to rethink how they inventory, govern, and decommission autonomous software.”
Saket Srivastava, Chief Information Officer, Asana
“By 2026, boards will ask the same questions about agents that they ask about people: who is allowed to do what, with which data and under whose supervision. A new risk category will emerge: shadow agents spun up without approval or oversight.”
5. Human–agent teaming and organizational change
Steven Webb, UK CTO, Capgemini
“Consumers value services that blend human empathy with the speed of AI. Businesses will focus on defining which tasks should be delegated to agents and how to measure their performance. Controlled environments will be crucial for testing human–machine collaboration safely.”
Tifenn Dano Kwan, Chief Marketing Officer, Amplitude
“In 2026, companies will no longer measure teams just by headcount. A new metric — agent count — will emerge as AI becomes core to how every function operates, sparking a fundamental shift in organizational design.”
Aditya Ganjam, Cofounder and CPO, Conviva
“The success of the agentic era will hinge not only on advances in AI but on the human culture that surrounds it. ‘Human-in-the-loop’ models will become the norm, embedding people at the center of decision-making and oversight.”
6. Specialization over general-purpose agents
Keith Zubchevich, President and CEO, Conviva
“2026 will be the year of specialized AI agents. One-size-fits-all agents will be replaced by purpose-built agents designed to drive specific outcomes. The winners will be companies building dozens of small, domain-specific agents aligned to clear goals.”
Ev Kontsevoy, CEO, Teleport
“The industry will create more granular terms to denote different types of AI agents, how they are deployed and whether they act on behalf of humans or have their own identity.”
7. Context, interoperability and platforms
Ken Exner, Chief Product Officer, Elastic
“The growth of agentic AI will hinge on accurate context engineering. Enterprises struggle with scattered, unstructured data, and failures often trace back to the inability to provide relevant context.”
Stuart Hubbard, Senior Director of AI, Zebra Technologies
“What will make cross-organization agents work are reliable ways to track context over time, memory layers and strong governance. 2026 will be the year foundations are put in place, not full autonomy.”
Dan Carpenter, CIO, Amplitude
“Differences between generative AI tools will shrink. Big platforms will rapidly increase agentic capabilities, driving a pivot back to platform AI to simplify and ‘agentify’ workflows using integrated datasets.”
8. Risk, experimentation and the “AI Wild West”
Thordis Thorsteins, Lead AI Data Scientist, Panaseer
“Agentic AI will pull businesses back into an experimental phase where risks rise as fast as opportunities. Autonomous systems amplify mistakes, making proactive governance, strict access controls and visibility essential to avoid a return to the AI Wild West.”
9. Sector-specific impact and sustainability
Yannick Chaze, Cofounder and CTO, Sweep
“AI is becoming a trusted copilot for sustainability professionals, reducing time spent on manual tasks and identifying emissions hotspots. From 2026, companies will standardize AI governance, with transparent human review and measurable environmental impact.”
Julius Černiauskas, CEO, Oxylabs
“Comprehensive agentic systems for public data gathering will drive down costs and democratize access. Multiagent systems can automate many small tasks, multiplying productivity as new tools enter the market.”