Brands are building toward a future where agentic AI manages the bulk of customer interactions and operational workflows, leading to increased employee productivity and creative outputs. Yet our global survey of 3,000 executives and practitioners for the Adobe 2026 AI and Digital Trends report finds that ambitions for AI-powered, real-time personalization at scale are running well ahead of organizational readiness.

Few have the data quality, harmonized profiles, or analytics frameworks required to scale their AI investments. While AI is already leading to improvements across key marketing workflows, teams are struggling to translate those benefits into quantifiable returns to leadership. Persistent misalignment on AI strategy between executives and day-to-day practitioners only compounds these challenges. To effectively scale AI-driven engagement, organizations must focus on closing three gaps simultaneously: data readiness, measurement infrastructure, and executive-practitioner alignment.

Brands are setting aggressive timelines for agentic AI, but today’s deployment progress points to a more uneven reality. Less than one-third of organizations have moved beyond the pilot stage into organization-wide or cross-functional deployment of agentic AI for customer support (30%) — an area that many expect to automate in the near future. Even fewer have deployed agentic AI to the same degree in marketing content creation and activation (22%) and personalization and recommendation (18%).

Despite their limited progress in scaling agentic AI, a majority of brands anticipate that the technology will directly handle a substantial share of all interactions across customer touchpoints. Within the next 18 months, more than three-quarters believe agentic AI will manage at least half of all customer support interactions. They hold similar expectations for agentic AI’s role in post-purchase support, customer sales and transactions, and conversational engagement (Figure 2). These aspirations about agentic AI’s growing role in customer engagement extend beyond scale — 42% of brands also plan to design distinctive AI agent personalities for different audiences, pointing to a broader reimagining of how customers experience AI-driven touchpoints.

Plans for agentic AI are equally bold when it comes to optimizing internal processes. Within the same timeframe, organizations expect agentic AI to be assisting employees with research and knowledge retrieval (69%), automating routine customer service tasks (63%), assisting marketing with campaign orchestration (52%), and creating content for marketing campaigns (44%).

These ambitions clearly indicate that brands are optimistic about the transformative benefits agentic AI can deliver. Nearly two-thirds (63%) believe it will free their organization to focus more on strategy and creative work, and 47% go further, saying businesses that fail to adopt agentic AI will become obsolete. Roughly one-third are already prioritizing agentic AI investment over more widely adopted approaches like generative AI. Yet as the gap between conviction and deployment suggests, moving agentic AI from experimentation to scale will require organizations to confront several operational barriers that ambition alone cannot resolve.

Despite the widespread perception of these benefits, demonstrating AI’s value is a persistent challenge. While organizations rank customer satisfaction and loyalty (e.g., Net Promotor Score, retention, churn) as leadership’s most important indicator of AI success — ahead of revenue growth and cost savings — more than half (56%) report that leadership ultimately evaluates AI outcomes purely through a financial lens. This disconnect is compounded by the fact that 52% of organizations struggle to demonstrate measurable returns on AI investments using any CX-related metrics, leaving the case difficult to make on either front.

Underlying both challenges is a lack of formal measurement infrastructure. Only 44% have implemented a framework for tracking the ROI and value of generative AI, and just 31% have done so for agentic AI. Nearly half (47%) have neither framework in place or do not know whether they do.