Adobe rebrands Experience Cloud as ‘CX Enterprise,’ goes all-in on AI agents

Adobe is replacing its Experience Cloud umbrella with a new flagship offering called Adobe CX Enterprise— an AI-first platform that merges creative and marketing capabilities under a single, agent-based architecture. The company plans to unveil CX Enterprise and more than a dozen related product announcements at Adobe Summit 2026, its annual marketer-focused event running April 20 to 22 at The Venetian in Las Vegas.

The rebrand signals a strategic shift from tool-centric marketing software to what Adobe describes as “goal-oriented, AI-first workflows.” CX Enterprise is organized around three pillars: Brand Visibility, Customer Engagement, and Content Supply Chain. Underpinning all three is a new Adobe AI Platform with two intelligence systems — Adobe Brand Intelligence, which governs brand consistency across AI-powered channels, and the CX Engagement Intelligence System, which handles optimization across audiences, channels, and customer journeys.

The move comes as the race to build agentic AI platforms intensifies across the enterprise software landscape. Salesforce aggressively pushed its Agentforce platform since its launch at Dreamforce in September 2024, where CEO Marc Benioff framed the event as a hard pivot to agents and declared the copilot era effectively over. Just days before Adobe Summit, Salesforce announced Headless 360 at its TrailblazerDX developer conference — an initiative that exposes its entire platform as APIs, MCP tools, and command-line interface (CLI) commands so AI agents can operate without a browser.

AI ‘coworkers’ and 10+ agents now generally available

The centerpiece of Adobe’s AI strategy is a new capability tier called “Coworkers” — persistent, self-learning agents with enterprise memory that can orchestrate multiple Adobe and third-party agents toward business goals.

Unlike one-shot agents that execute a single task, Coworkers run continuously, learn from outcomes, and can be triggered by signals or schedules. Adobe says the Coworker concept targets all CX personas — developers, marketing ops, and marketers — not just technical users.

More than 10 purpose-built AI agents previewed at Summit 2025 are now in production, including agents for site optimization, data insights, audience creation, journey orchestration, experimentation, LLM optimization, and content optimization. Adobe says 1,770-plus customers are already entitled to use them through a new credit-based pricing model. At Summit 2026, Adobe is positioning these agents as the foundation layer of CX Enterprise, with the new Coworker tier sitting above them as a longer-running, goal-oriented orchestration capability.

The governance question

Enterprise customers are not universally enthusiastic about autonomous AI agents. Unpredictability and governance are emerging as top concerns, with significant pushback reported around the levels of agentic autonomy organizations are being asked to accept.

Adobe says it is incorporating two distinct levels of human oversight depending on the use case. Human-in-the-Loop means a human is actively involved in the process — reviewing, approving, or redirecting the agent’s work before it proceeds. Human-on-the-Loop means the agent operates autonomously within preset guardrails, with humans monitoring outputs and able to intervene, but not required to approve each step.

Brand Concierge — the consumer-facing conversational agent — will run Human-on-the-Loop, with guardrails, policies, and quality training baked in, plus monitoring for ongoing fidelity to organizational goals. Design-time activities like campaign planning will use Human-in-the-Loop, where Adobe says there’s more opportunity for ideation and interrogation — in other words, use cases where human judgment is part of the creative process, not just a safety check.

On quality management, Adobe says it has invested in internal tooling and external partnerships to validate agent outputs, including using AI models to evaluate other AI models — a technique known as LLM-as-a-judge. The company attributes overall quality improvements to advances in the underlying models themselves, as well as a validation step built into the agent workflow that catches and corrects errors before they reach end users.

Adobe’s own AI & Digital Trends Study from March 2026 underscores the challenge: 75% of organizations cite data integration and quality as their top AI implementation challenge, 71% cite talent gaps, and 68% cite unclear ROI.

Open agentic ecosystem: MCP, A2A, and partner integrations

Adobe is betting heavily on interoperability. The platform now supports Model Context Protocol (MCP) across its products and offers reference architectures for Microsoft Copilot, ChatGPT Enterprise, Claude Cowork, and Gemini Enterprise. Adobe CX Skills — pre-built capabilities from across the platform — are available inside those partner AI environments. The company offers three deployment models: full-stack Adobe, Adobe-orchestrated within partner UIs, or bring-your-own orchestration.

Adobe is expanding the role of Workfront — its project management and workflow platform — beyond its traditional function as an internal “system of work” for creative and marketing teams. Under the new Agency System of Record model, Workfront becomes a shared operational layer that connects brands directly with their agency partners, giving both sides visibility into the same workflows, assets, and project data.

The five largest agency holding companies — WPP, Publicis, Omnicom, Dentsu, and Havas — are integrated into this layer at launch, meaning creative and campaign work moving between a brand and any of those agencies would run through a common operational infrastructure rather than the fragmented mix of email threads, shared drives, and separate project tools that typically govern brand-agency collaboration today.

When asked how agents balance competing goals and organizational constraints, Adobe said agents will operate within a defined enterprise context layer that encodes organizational goals, policies, and guardrails. Customers can feed both hard constraints — such as compliance policies — and higher-level business objectives into that layer, allowing agents to prioritize and make decisions accordingly.

GenStudio expands from one product to three

Adobe GenStudio, which launched six quarters ago as a single product for performance marketing, is expanding to three modules:

• GenStudio for Performance Marketing gets a major update with an omnichannel insights dashboard, automated creative element tagging, proactive optimization recommendations, and multi-model AI support (Firefly Image Model 5 plus third-party and custom brand-trained models). Adobe is also adding two new activation channels: ChatGPT, where brands can now publish and activate campaigns directly, and connected TV, enabling the creation of CTV ad formats within the platform.
• GenStudio for Content Marketing is a new product entering early access at Summit. It atomizes long-form content — white papers, thought leadership, research — into social clips, short videos, posts, and emails. Adobe says the tool reduces timelines from four to six weeks down to minutes. It includes LinkedIn and Meta scheduling with performance attribution built in.
• GenStudio for Commerce Media targets media networks and long-tail advertisers with self-service ad creation, co-brand compliance checks, and an embedded SDK for integration into existing self-serve platforms.

On the creative side, Adobe’s most direct consumer-facing competition is increasingly with Canva, which has accelerated its AI roadmap aggressively since Adobe’s failed attempt to acquire Figma — a $20 billion deal blocked by EU and U.K. regulators in December 2023. Both Adobe and Canva have since launched AI-powered creative workflows: Adobe through its Firefly AI Assistant in Creative Cloud, and Canva through its AI design model platform.

Real-time CDP: Unstructured data and new agency and media offerings

Adobe Real-Time CDP is adding unstructured data support — call center logs, chat transcripts, and video interactions — to customer profiles via vector embeddings. The goal is to connect customer intent with behavioral data, giving AI agents richer context for personalization.

Three new CDP Collaboration offerings are launching:

• CDP Collaboration for Agencies is a dedicated product for agencies to manage data collaboration on behalf of brand clients; Epsilon is the first integration partner, using Core ID for identity resolution, with WPP next.
• CDP Collaboration for Media Networks provides a white-labeled portal for advertiser and agency onboarding with clean room-based audience planning and measurement.
• Universal Collaboration, now generally available, lets any organization collaborate without license requirements, regardless of data warehouse, CDP, or identity vendor.

New integrations include TikTok (campaign measurement and planning), The Trade Desk (first-party data enrichment with exposure signals), Amazon Ads/Amazon Marketing Cloud (moving to GA), and data partnerships with Acxiom and Experian.

Adobe has also significantly upgraded the platform’s ability to handle traffic spikes during high-stakes moments — think Super Bowl campaigns, tax season, or major product launches. The system can now scale dramatically beyond standard capacity on demand, and the time it takes to refresh customer data has been cut from three days to 14 seconds.

Marketo Engage gets agentic AI and an MCP server

Marketo Engage is getting a conversational AI layer with purpose-built agents for program creation, campaign validation, smart list building, lead import enrichment, and data normalization. A new Marketo MCP Server positions Marketo as a hub in agentic martech stacks, enabling agent-to-agent communication.

The Marketo MCP server enters a space where third parties like Inflection.io, Zapier, and CData already offer Marketo MCP servers, suggesting enterprise demand for agentic marketing automation was ahead of Adobe’s own development timeline. The product also gets a refreshed UI with Adobe’s current design language and a built-in AI assistant for product knowledge.

AJO B2B Edition: New entry-level tier and sales qualifier

Adobe Journey Optimizer B2B Edition is splitting into two tiers. A new AJO B2B Edition Prime, available in Q2 2026, uses Marketo Engage data as a lower-entry point. The existing product becomes AJO B2B Edition Ultimate, creating a maturity path: Marketo Engage to AJO B2B Prime to AJO B2B Ultimate (with Real-Time CDP B2B).

AJO B2B also adds Semantic AI Decisioning for next-journey and next-step recommendations, plus a new Sales Qualifier — an AI-first tool for business development reps that handles qualification, personalized outreach, and autonomous next-step decisioning.

Pricing: Credit-based and ‘value-based’ models

Adobe is moving toward centralized, credit-based pricing for AI agents. New AI-first apps are priced on what Adobe calls “value-based metrics.” A more detailed practitioner dashboard breaking down usage patterns and prompts used by teams is currently in beta.

What it means for marketers

Adobe, Salesforce, and HubSpot are converging on the same strategic vision: platforms that operate as infrastructure for AI agents rather than interfaces for human users. All three are now positioning automation layers as the primary way work gets done, with Salesforce making this explicit with Headless 360, and HubSpot pursuing similar territory with its Breeze AI agents across the mid-market.

The practical question is how quickly this translates from keynote demos to production workflows. Adobe’s own research shows the headwinds are real — three-quarters of organizations still struggle with data quality, and more than two-thirds can’t demonstrate clear ROI from AI implementations.

Concern around autonomous agent governance adds another layer of uncertainty. The GenStudio expansion and CDP Collaboration offerings are arguably the most immediately actionable announcements — they address specific, well-understood pain points around content production velocity and cross-organization data collaboration that don’t require enterprises to fully embrace autonomous AI.

Summit runs April 20 to 22 with approximately 14,000 expected in-person attendees and more than 250 sessions.