How enterprises are building AI agents in 2026

AI agents have evolved from experimental tools into a critical component of the modern enterprise technology stack. But how are organizations actually using them today—and where is the real value emerging?
In collaboration with research firm Material, we surveyed more than 500 technical leaders across industries and company sizes to understand how AI agents are being deployed now and how enterprises plan to expand their use in the coming years.
The results reveal a clear shift: organizations are moving beyond basic task automation toward sophisticated, multi-step workflows that span teams, systems, and business functions.

What the Data Reveals

AI agent adoption is accelerating in both scope and complexity. Today, 57% of organizations are using agents for multi-stage workflows, with 16% already running cross-functional processes across multiple teams. Looking ahead to 2026, 81% plan to pursue more advanced use cases, including 39% developing multi-step workflows and 29% deploying agents across departments.

Software development remains the leading entry point. Nearly 90% of organizations use AI to support development activities, and 86% deploy agents directly in production code. Teams report significant time savings across the entire development lifecycle—planning and ideation (58%), code generation (59%), documentation (59%), and testing and review (59%).

However, the impact extends well beyond engineering. Data analysis and report generation (60%) and internal process automation (48%) rank among the highest-value use cases. Over the next year, 56% of organizations plan to expand agent usage into research and reporting.

Most notably, 80% of respondents report measurable economic returns from their AI agent investments—demonstrating that these tools are already delivering tangible business value.

What This Looks Like in Practice

The organizations achieving the strongest results are no longer treating AI agents as experiments—they are embedding them into core infrastructure.

• Thomson Reuters uses AI to power CoCounsel, enabling legal professionals to search 150 years of case law and tap into thousands of domain experts in minutes rather than hours.
• eSentire, a cybersecurity firm, reduced expert threat analysis time from five hours to just seven minutes, with AI outputs aligning with senior analyst decisions 95% of the time.
• In healthcare, Doctolib deployed AI-driven development tools across its engineering teams, replacing legacy testing systems in hours instead of weeks and accelerating feature delivery by 40%.
• In retail, L’Oréal achieved 99.9% accuracy in conversational analytics, allowing more than 44,000 monthly users to query data directly rather than waiting on custom dashboards.

The Path Forward

For enterprise leaders, the key question in 2026 won’t be whether to adopt AI agents—but how to scale them effectively. The data highlights three primary challenges: integration with existing systems (46%), data access and quality (42%), and change management (39%).

At the same time, the organizational impact is undeniable. Nine out of ten leaders report that AI agents are reshaping how teams work—freeing employees to focus more on strategic thinking, relationship-building, and skill development instead of routine execution.

Successfully scaling agents requires purpose-built infrastructure: models optimized for enterprise workflows, frameworks such as agent development SDKs, and tools that enable teams to move from prototype to production quickly and securely.

While software development has been the proving ground for AI agents, it’s only the beginning. As agents expand into research, customer support, financial planning, and supply chain operations, organizations that build expertise now will be best positioned to capture outsized value as the technology continues to mature.

celawebmaster