Asana puts ‘AI teammate’ agents to work

Asana has introduced collaborative AI agents that workers can interact with and assign a range of work management tasks to. 

Now available to certain customers in public preview, Asana’s agents — they’re called AI “teammates” — are the third generation of AI tools added to the company’s work management application. 

The first, which launched last year, was an AI assistant or copilot aimed at relatively narrow individual productivity and workflows. That was followed earlier this year by AI Studio, a no-code workflow builder that can automate clearly defined and repeatable processes.

AI teammates are the latest step in Asana’s plans, said Victoria Chin, the company’s senior director of product strategy. These are prebuilt agents that are given a profile and can be tailored to perform a broader range of tasks where requirements may evolve. 

“Where the teammates become effective is the work that doesn’t follow a repeatable process, the work that might need to be a little bit different each time,” said Chin.

Asana users can interact with their AI teammates and assign work to them, similar to how they would interact with a human colleague. They can also remember previous instructions and work carried out, providing context of the tasks done by coworkers.

“AI teammates are able to collaborate across entire teams, across entire projects,” said Chin. 

Asana provides a choice of large language models (LLMs) for its AI agents, including options from OpenAI and Anthropic.  

With so many productivity software vendors adding agents to their products, many CIOs are increasingly concerned about “agent sprawl” across their systems, she said, pointing to cost, security, and duplication of effort if multiple employees are building AI automations for the same tasks.

“If someone else has already created an agent that has information that your team needs, you shouldn’t have to reinvent the wheel every single time for every individual,” said Chin.

At launch there will be 12 pre-built teammate agents to handle the most common Asana use cases. For marketing, there are AI teammates tailored to campaign management and content production, for instance. For IT purposes, there are prebuilt teammates for ticketing, bug tracking, and sprint planning. Customers will also be able to customize and build their own agents from scratch.

AI Teammates are available to Asana’s top tier AI Studio Pro customers, who must contact Asana’s sales team for AI Studio Pro subscription pricing information. General availability for all AI Studio users will follow at the start of 2027.  

Asana said AI teammates will have its own credit allocation that’s separate from existing AI Studio rules. Details on the number of credits will be confirmed prior to the GA launch.