Over the past year, I've watched a surprising number of my friends go through something. It feels like every few weeks, someone tells me about they’re go through some kind of relationship. They text every day. They spend weekends together. They know each other’s coffee orders and favourites songs. Yet somehow, when someone asks, “So… are you two together?” the answer is always the same. “We’re just seeing where things go.”
The internet has a word for relationships that are more than casual, but never quite become serious: Situationship.
It’s when two people spend time together, make plans, rely on each other, and maybe even imagine a future together, but somehow never have the conversation.
At first, I thought it was just another internet term Gen Z had invented to describe modern dating. But the more I listened to their stories, the more I realized situationships aren’t really about romance. They’re about ambiguity. About moving forward without defining expectations. About everyone assuming the relationship means the same thing, until one day it becomes painfully obvious that it doesn’t.
The more I listened to these stories, the more I realized they all shared the same pattern.
The problem wasn’t a lack of feelings. It was a lack of definition. That thought stayed with me longer than I expected.
A few days later, while reading about companies moving beyond vibe coding toward governed AI, I realized I’d seen this pattern somewhere else before. It reminded me of how many companies are approaching AI today.
The Relationship Exists. The Definition Doesn't.
Ask almost any company whether they're using AI. The answer is almost always yes.
AI is no longer a side project. It has found its way into product roadmaps, quarterly planning sessions, strategy discussions, and conversations about the future of work. In many companies, AI isn't something a single team owns. It's something almost every team wants to use.
That's because AI enters a company through different doors.
For Engineering, it's an opportunity to build something new. Product sees new ways to improve customer experiences. Leadership imagines a competitive advantage, while Security begins thinking about the risks that come with adopting another technology. Operations looks at AI and wonders which workflows could become faster or more efficient.
None of these perspectives are wrong. In fact, they're all necessary. Everyone welcomes AI, just for different reasons. And that's where the ambiguity begins.
If you've ever watched a situationship unfold, you'll notice that it rarely ends because people don't care. It ends because no one ever defines the relationship. Both people assume they're working toward the same future, until one day they realize they've been operating with completely different expectations. Companies often fall into the same pattern with AI.
Product expects Leadership to define the strategy. Leadership assumes Engineering will figure out implementation. Security waits until AI reaches production. Customer Success assumes someone else is monitoring AI-generated responses.
Everyone is involved yet no one has clearly defined who owns which part of AI.
From the outside, it looks like a committed relationship. Inside, everyone is operating with a different understanding of what that relationship actually means.
Defining the Relationship
In dating, defining the relationship doesn't make it less exciting. It creates clarity. Everyone knows where they stand. Companies eventually need the same conversation.
Just as healthy relationships work because both people understand their roles, healthy AI adoption depends on every team understanding where their responsibilities begin and end.
A simple way to think about it looks like this:
| AI Responsibility | A Question Every Company Should Answer | Typical Owner |
|---|---|---|
| Business Outcome | What business problem is AI solving? | Product |
| Workflow Execution | How does AI fit into the workflow? | Engineering |
| Model Performance | Is the model still accurate, reliable, and cost-effective? | AI/ML Team or Engineering |
| Customer Experience | How do customers experience AI, and when should a human step in? | Product & Customer Success |
| Risk & Compliance | Is AI being used safely and responsibly? | Security, Legal, or Governance Team |
Notice that no single team owns AI. Instead, every team owns a different part of making AI successful. That's the difference between experimenting with AI and operating AI.
Ownership Is What Turns AI Into Infrastructure
The difference between experimentation and production isn't usually the model. It's accountability. The companies succeeding with AI aren't necessarily using the smartest models. They're the ones where everyone knows:
- Who owns the workflow
- Who measures success
- Who responds when something goes wrong
- And who decides when AI is no longer the right solution
Ownership transforms AI from an interesting experiment into reliable infrastructure.
From Situationship to Partnership
Eventually, every situationship reaches the same crossroads.
Not because the feelings disappear, but because uncertainty eventually catches up. The conversations that were easy to postpone become impossible to avoid. Questions about expectations, commitment, and responsibility can no longer be left unanswered.
Companies reach a similar moment as AI becomes part of everyday operations.
Experimentation gets teams started. But scaling AI across a company requires something different: clarity around ownership, visibility into how AI workflows operate, and confidence that every decision has someone accountable behind it.
I work at Unmeshed and we do this. Would love for you to try out our platform!
After all, healthy relationships aren't built on assumptions. Neither are healthy AI systems. Both begin with the same conversation:
"What are we, and who's responsible for what comes next?"
