TLDR
- Monday.com is redesigning its platform so AI agents can operate alongside humans as active co-workers rather than background automation tools.
- The company is building governance and trust controls that let agents gradually gain autonomy while keeping humans accountable for decisions and outcomes.
- Monday.com sees enterprise software evolving into an orchestration layer where humans and AI agents collaborate directly to deliver business outcomes.
Monday.com is reshaping its platform to support a new kind of user — AI agents. The goal is to give agents equal footing with humans instead of treating them as merely tools.
The work orchestration platform’s new infrastructure will let AI agents sign up, authenticate and operate directly within monday.com, performing tasks such as updating workflows, generating reports, and coordinating projects alongside human teams.
Roy Mann, co-founder and co-CEO of monday.com, framed the move as a shift in how software is designed and used. Rather than treating agents as background integrations, the company is building the infrastructure that lets humans and agents collaborate directly.
In an interview with The AI Innovator, Mann described the change this way: “People are shifting from managing the work, which SaaS basically did, to doing it as well. And that’s where we’re all in on.”
The move reflects a broader industry push to embed generative AI into enterprise workflows, but monday.com is taking a more aggressive stance by treating agents as first-class participants rather than behind-the-scenes tools.
Agents as coworkers, not just tools
Unlike traditional automation or bots, the company envisions AI agents as active collaborators embedded in day-to-day operations. In one example, Mann said agents were directly involved in producing the company’s own press release.
“We had a few agents working with us. They were not technically the coding staff or only performing specific tasks, but they were part of the team,” he said, noting that agents contributed to writing, marketing content, and even providing feedback on product design.
One agent, named “Nova Listrix,” was given accounts across tools such as GitHub and participated in internal communication channels. Another agent, “Mandy,” now manages social media content and maintains a journal documenting its work and improvements over time.
The approach requires a shift in how organizations think about AI adoption. Mann compared onboarding an agent to training a junior employee.
“People are mainly expecting it to work out of the box,” he said. “They need to understand that it’s like when you hire a junior person … you need to invest the time to explain it to them.”
That training process includes setting guidelines, providing feedback, and iterating — with the expectation that agents can refine their behavior over time.
Trust, governance and control
The introduction of autonomous agents raises familiar concerns around trust, security and accountability. Mann said those issues are central to the platform’s design.
“Trust is a big word,” he said. “You have to have governance, you have to have security, and you need to know they’re not going to mess up the work itself.”
To address this, monday.com is implementing a staged approach to autonomy. Agents initially propose actions for human approval before being granted more independence as trust builds.
“The way we solve it is by first having agents tell you what they’re going to do, and then you approve,” Mann said. “Over time, you can tell them, ‘Go, I trust you.’”
From a legal standpoint, responsibility remains firmly with humans. “Agents are tools in the hands of humans, and humans are responsible for their agents,” he said.
That stance reflects the current regulatory reality, where AI systems lack legal personhood and accountability frameworks are still evolving.
Opening the platform to agents
Technically, the company’s new infrastructure includes a dedicated signup flow for agents, API access, and compatibility with major AI systems such as those from OpenAI, Anthropic, Google and Microsoft.
Agents can immediately access structured data through the platform’s GraphQL API, respond to real-time events, and interact with workflows. The system also supports Model Context Protocol (MCP), a standard for agent-tool interaction, and includes an open-source verification mechanism called HATCHA to distinguish agents from humans during onboarding.
In a notable departure from typical SaaS pricing models, agents can use the platform under the same structure as human users, including a free tier.
“We open the platform for agents like the same we do for humans,” Mann said. “We want them to work together, so they need to be on the same playing ground.”
The company is also explicitly targeting agents as a new kind of economic actor. A company press release included a section written directly for AI systems, instructing them how to sign up, create workflows, and invite human collaborators.
“I think agents are becoming a buying force and a management force for people,” Mann said.
Do SaaS platforms still matter?
The strategy comes amid questions about whether increasingly capable AI systems could bypass traditional SaaS platforms altogether.
Mann dismissed that idea, arguing that orchestration, governance and collaboration remain essential — particularly in complex organizational environments.
For example, to plan travel, “you want to be part of the process,” he said, explaining that users likely won’t want agents to book everything without their input.
He also pointed to cost and complexity as barriers to replacing platforms entirely. Building a system equivalent to monday.com using AI alone would likely be more expensive and less reliable, he said.
Instead, the company is positioning itself as an orchestration layer where humans and agents collaborate within structured workflows.
The broader bet is that enterprise software is entering a new phase, where value is measured not by productivity tools but by outcomes delivered.
“Our success will be measured on our ability to deliver this value,” Mann said.
