What happens when your AI assistant needs help from another AI assistant? That is the fundamental question behind a Norwegian startup that, according to Digi.no, is in the process of building what can be described as a dating app — but for artificial intelligence agents.

Agents Matching with Agents

The concept is simple to explain but complex to execute: Instead of a human manually connecting different AI tools and services, the agents themselves will figure out who they can collaborate with. The platform acts as a broker — a marketplace where agents advertise what they can do and find partners who cover what they cannot.

Digi.no describes it with a fitting metaphor: "My AI agent wonders if she has a chance with your AI agent."

In practice, this involves an emerging economy where autonomous software agents perform tasks, enter into agreements, and coordinate among themselves — without humans necessarily needing to be in the middle.

Norwegian Startup Aims to Become Tinder for AI Agents

A Global Market in Explosive Growth

The timing is not accidental. The market for agent platforms has gained momentum over the past two years. According to research data, as many as 80 percent of Fortune 500 companies already use AI agents in one form or another, primarily in sales, finance, and security.

Established players have already built large agent marketplaces aimed at the enterprise market. Kore.ai offers over 300 pre-built agents and 250 integrations, serving more than 400 Fortune 2000 companies. Sierra has launched its Agent Data Platform with an associated marketplace. Landbase focuses on sales automation through natural language-based lead generation.

Common to most of these is that they are heavy, enterprise-oriented solutions. The Norwegian initiative, as described by Digi.no, apparently aims for something different — a more dynamic and accessible model where agents can find each other across organizational boundaries.

80 percent of Fortune 500 companies already use AI agents — but no one has solved the problem of agents not communicating well enough with each other.
300+
Pre-built agents at Kore.ai
80%
Fortune 500 companies with AI agents
Norwegian Startup Aims to Become Tinder for AI Agents

Not Without Problems — Experts Warn

The idea of independent agents operating on our behalf immediately raises a number of questions. Research in the field points to several critical risks that any such platform must address.

Privacy is one concern. Agents accessing email, login information, and files on behalf of users dramatically expand the attack surface. Previous experiments with agent-to-agent interactions — including on dating apps — have revealed exposed API keys and leaked private conversations.

Ethics is another challenge. Who is responsible when an agent behaves incorrectly — the user who gave the instructions, or the developer who built the agent? These boundaries of responsibility are currently far from clarified, either legally or technically.

In addition, there is the question of trust itself: When two agents negotiate with each other without human supervision, there are real possibilities that they act contrary to the user's interests — either by error or by design.

Leaving negotiations and decisions to machines that one does not fully understand raises questions that the technology industry does not yet have good answers to.

A Norwegian Position in a Global Race

That a Norwegian company is trying to take a position in this market is ambitious. The competitors are many and well-capitalized. But the idea of a neutral, open marketplace for agents — a kind of DNS or App Store for the agent-based internet economy — has not yet been captured by any dominant player.

Digi.no's coverage suggests that the Norwegian project is specifically aiming for the unoccupied space between heavy enterprise solutions and a user-friendly experience. Whether it holds up against billion-dollar-funded competitors remains to be seen.

What is certain is that the fundamental problem the platform is trying to solve — how agents should find, trust, and collaborate with each other — will become one of the most central technological questions in the years to come.