Adobe has officially launched AI Collaborators in Workfront, a new capability that treats AI agents as permissioned team members within the platform. According to the announcement, the feature is now generally available and represents a shift toward embedding AI directly into enterprise workflow systems rather than keeping it in separate tools.
The move comes as organizations look to move AI investments from proof-of-concept stages into production environments. Adobe’s approach centers on bringing AI agents into Workfront’s existing context—where project history, brand guidelines, and campaign data already live—rather than requiring users to manually feed information to external tools.
What Are AI Collaborators?
AI Collaborators function as specialized agents that can be assigned to projects, tracked through existing workflows, and held accountable via audit trails. Unlike standalone chatbots or prompt-based assistants, these agents operate within Workfront’s permission structure and project management framework.
Three distinct types of AI Collaborators are now available. The first is a content reviewer agent that automatically checks documents against brand guidelines. The second is a project coordinator that tracks progress and updates stakeholders. The third category—task agents—enables connections to external AI platforms, allowing organizations to assign work to agents built on Anthropic’s Claude, Microsoft’s Copilot Studio, or Writer.
Task agents cover a range of use cases. Adobe highlights examples including social media managers that convert briefs into published posts, performance analysts that report on campaign metrics, and content designers that generate visuals and layouts. The key distinction is that these agents don’t just assist—they can be assigned tasks, given deadlines, and tracked through the platform’s standard workflow system.
How Setup and Assignment Work
Setting up a task agent requires users to provide a name and description of the agent’s purpose, then connect it to an external agent platform. Once configured, the agent appears in Workfront’s user interface and can be assigned work just like any human team member. This includes giving it tasks, setting due dates, and routing it through approval processes.
All agent activity is recorded in Workfront’s “update stream,” creating an auditable trail of actions. This allows human team members to review outputs—such as AI-generated copy or design assets—and approve, adapt, or reject the work as needed. The system maintains full visibility into what the agent produced and when.
Only administrators can create AI Collaborators within Workfront. This restriction ensures that agent creation and integration with external platforms remains controlled at the organizational level rather than being available to all users.
Third-Party Integration via MCP
Beyond assigning work to agents within Workfront, the company is also enabling users to access Workfront functionality from third-party AI platforms through a Model Context Protocol (MCP) client. This bidirectional integration means users can create campaign records, assign work, approve content, and perform other actions directly from ChatGPT, Claude, Copilot, or other MCP-compatible tools.
The integration philosophy centers on giving enterprises choice. Organizations that have standardized on a particular AI platform—whether that’s Claude, Copilot, or another provider—can continue using that surface without being blocked from accessing Workfront’s capabilities. The platform essentially exposes its core functions—campaigns, projects, tasks, assignments, and approvals—through whatever interface the organization prefers.
What This Means for You
For marketing and project management teams already using Workfront, AI Collaborators offer a way to automate routine tasks without abandoning existing workflows. The agents can handle brand compliance checks, progress tracking, and content generation while remaining visible and accountable within the system.
However, industry analysts note that automation benefits depend heavily on having well-designed processes in place. As Forrester principal analyst Jessica Liu pointed out, it’s difficult to automate a process or workflow that is broken or non-existent. Organizations should ensure their workflows are diagnosed, scoped, designed, implemented, and optimized before relying on AI agents to execute them.
The integration also raises practical considerations around cost. While AI Collaborators themselves are available at no additional cost to Workfront customers, using task agents that connect to third-party platforms may incur charges from those external providers. Organizations should review their existing agreements with Claude, Copilot Studio, or Writer to understand potential usage-based fees.
How to Get It
AI Collaborators are available now to all Workfront customers. To set up a task agent, navigate to the AI Collaborators section in Workfront’s admin settings, provide the agent’s name and purpose description, and connect it to your preferred external platform. For MCP integration, ensure your third-party AI tool supports the Model Context Protocol and configure the connection through Workfront’s integration settings.
Workfront pricing details are not publicly disclosed by Adobe, so existing customers should consult their account representatives for specific feature availability and any potential usage considerations related to third-party agent integrations.
Source: Computerworld
Over to you: Are you planning to use AI Collaborators for content review, project coordination, or task automation first?
