You’re reading an issue of “The AI Economy,” my newsletter exploring the forces shaping the AI era—tracking how AI is rewriting business, work, technology, and culture. Subscribe to get expert insights and curated updates delivered straight to your inbox.
Asana has spent over 18 years telling companies how to manage work. Now, it wants to tell them how to manage AI. On Thursday, the company unveiled its Agentic Work Management Platform, a system designed to align humans and agents around the same plan, context, and governance. Asana calls it the “most significant product evolution” in its history, and it comes with three parts: a personal chief-of-staff agent called Dash, a new generation of AI Teammates, and three vertical applications targeting IT, software development, and professional services.
While many enterprise software vendors envision AI agents as being autonomous workers and humans as supervisors, Asana’s approach is more constrained and structured. It reports that although three-quarters of knowledge workers use AI in their jobs, only five percent of organizations report meaningful gains in productivity. A key reason why agents fail in the enterprise: A lack of organizational memory with no understanding of who does what, by when, or why, explained Chief Product Officer Arnab Bose. Asana’s argument is that it has spent years building exactly that infrastructure, even before agents existed to use it.
“It enables us to create this system where human beings and AI agents can work side-by-side in the flow of business,” he told The AI Economy in an interview, arguing it delivers better performance and quality outcomes compared to agents operating through chat-based tools like a Copilot sidebar. “We created the collaborative work management category,” Bose said. “Now we’re creating a new category: agentic work management.”
The ‘Easy Button’ Enterprises Are Looking For?
Asana isn’t the first to talk about how AI agents need more context to be successful. Glean, Slack, ServiceNow, and Atlassian are among those sounding the alarm lately. But where Asana has a structural edge is in four key areas, it argues.
The first is the Work Graph itself, which functions as a true context graph, tracking not just what work exists but who owns it, when it’s due, and why it matters. The second is its multiplayer interaction model, which makes all agent activity visible to the entire team, not just to the person who invoked it. Third is shared memory—as AI Teammates interact with multiple workers, they retain what they learn, so a new team member doesn’t have to retrain the agent from scratch. Finally, every agent action is logged in a full audit trail, including what was accessed, who owns the agent, and how much it costs to run.
Together, these four capabilities can help bestow the organizational memory that AI agents need to transform a productivity experiment into a reliable system of record. “There’s tremendous pressure on all businesses worldwide to become AI native,” Bose said. “Everybody feels like the technology space is moving super fast, that they’re missing out on an ability to accelerate their business, accelerate work, and gain an edge.” But while technology continues to improve, there has yet to be a so-called “easy button” for the agentic enterprise, he contended.
Asana thinks it has cracked the formula and has the “easy button” companies need.
“Not only are we giving you this easy button to identify those processes, the AI agents are making it easier than ever before to create the projects, keep them updated, and create this compounding benefit for your company,” Bose said. “Don’t think of Asana as this work management tool where you have to track and log work. Think of it as this operating system that gives you the easy button to make your enterprise truly agentic.”
Asana Dash
Asana Dash is a personal AI assistant built for the individual worker. It’s distinct from the company’s AI Teammates, which are multiplayer by design. Where Teammates operate as collaborative agents within a team’s workflow, Dash is yours alone. It understands your goals, priorities, and pending work, handling what it can and surfacing what needs your attention.
“We are explicitly not calling [Dash] a Teammate because we want people to believe it’s their assistant—it’s your Dash,” Bose said. “It has all the capabilities of a Teammate, and it can invoke Teammates for you. It can update projects, tasks, portfolios—things like that. But it really is for you.”
Asana subscribes to a vision of digital labor in which AI agents handle the busy work, freeing human workers to focus on higher-order decisions and judgment calls. But as the pace of work accelerates, every worker increasingly becomes a manager of AI agents, and managing agents while staying in the flow of work requires its own layer of support. That’s where Asana Dash fits in.
Bose shared that he’s using Dash to generate a morning brief for himself, scanning his email inbox for unread emails over the past 24 hours, unread Slack messages from select channels, and updates and activities on critical projects he’s tracking. Each morning, he receives a notification with a summary, action items, and required approvals. “All of these things are just helping me get going with my game much faster,” Bose said.
New AI Teammates
Asana’s AI Teammates were introduced two years ago as customizable workflows that can take action when needed to get jobs done. They’re also powered by the company’s Work Graph data model. In September 2025, Asana upgraded its Teammates so they function as actual coworkers and operate actively rather than passively. They only recently became generally available in March.
The next-generation Teammates announced on Thursday are more capable, more connected, and easier to adopt. Asana has redesigned the chat interface for interacting with these agents and added in-product recommendations that surface the right Teammate for the job. A new Skills library lets teams extend Teammates with additional capabilities tailored to their specific workflows, capturing repeatable work patterns. And new integrations with Gmail, Outlook, Slack, HubSpot, Figma, and Canva allow Teammates to handle multi-step work across apps workers use daily.
Asana is also launching industry-specific AI Teammates for manufacturing, retail, and other sectors, expanding beyond its existing marketing, IT, and operations agents. Bose said one such Teammate is designed to address launch planning: “People are using Asana to look at a marketing or product launch. There are many different stakeholders publishing updates, and you have to go through as a human being and look at tech space status updates to glean if a project is on track or not. These launch planner Teammates can automate all of that process.” In other words, it’ll proactively scan through status updates and flag any risk or remediation steps before guiding the worker through the process. The Teammate will then be responsible for chasing down stakeholders for updates, analyzing them, and flagging any additional risks.
“The human being is elevated to being the tastemaker, where they can evaluate if the quality of the remediation is what they want, if they want to run a different play, [or] how they want to do stakeholder management with other human beings,” Bose explained.
Apps for the Human-Agent Operating System
Beyond the core platform, Asana is making its first foray into verticalized applications, packaging its agentic capabilities for the teams whose work most depends on coordinated execution.
Asana Service Management brings together support ticketing and project management and is aimed at IT, HR, facilities, and other service workers. The app features a self-learning knowledge base that can improve case deflection over time. The company claimed that if another team needs to get involved, Asana can move from a ticket to a project without sacrificing context.
The move puts Asana in direct competition with Atlassian, ServiceNow, and Salesforce, and Bose doesn’t shy away from it. “We are seeing a tremendous amount of interest from our customer base in replacing their existing IT ticket management tools with something more modern,” he said. Many customers, Bose added, already run Asana alongside ITSM tools, using it for project management. “We are combining those two things, and in the cases where people buy us, they will be displacing an existing ticketing product.”
Command by Asana is a software development system designed to turn ideas into outcomes. It is not a coding agent or a standalone command-line interface (CLI), but rather a framework orchestrating the entire development lifecycle, from generating product requirements documents (PRDs) and well-documented tickets to coordinating with coding agents, tracking pull requests, and keeping engineering and product leaders current on velocity, risk, and delivery timelines.
Asana Client Management addresses new client onboarding by coordinating all deliverables and requested approvals. It features six pre-built specialized AI Teammates designed for agency work, such as client intake, project staging, producing deliverables, and communicating status updates.
Asana 2.0 Coming Later This Summer
Bose described the Agentic Work Management Platform as Asana 2.0, elevating collaborative work management to a mode in which humans and agents work together from a shared system. The full suite is expected to be available later this summer, at which point every Asana customer will automatically receive a baseline allocation of AI Teammate executions and Dash access, no procurement required.
Thursday’s announcement caps off a busy week for Asana. Earlier, the company revealed it had acquired the no-code, agentic development platform startup StackAI for $75 million. However, Bose noted that today’s announcement didn’t involve StackAI at all, at least for now. Once Asana’s Agentic Work Management Platform launches, the company plans to let customers build agents in StackAI and then integrate them with AI Teammates within Asana.
“That agent that is doing multi-party orchestration will now get the benefits of organizational context with the Work Graph, shared memory, a multiplayer experience for human beings, and the governance and activity tracking that we have within Asana,” he said.
There’s a lot riding on this bet. Asana has been heavily impacted by generative AI, as have most SaaS companies—SaaSpocalypse, anyone?—with its market valuation sliced nearly in half. Whether its pitch as the operating system for human-agent teams wins over more customers remains an open question. Bose said that Asana 2.0 is about “reinventing our existing $800 million customer base.” Still, it’s hardly a unique position in a market where ServiceNow, Salesforce, Microsoft, and others have all staked their own claims to owning the agentic enterprise.
