Nobody Owns the Loop
Cognizant published research this week that landed at the Fortune COO Summit.
Ninety-three percent of jobs are already impacted by AI. Thirty percent are facing what their researchers call “existential change” — not task substitution, but structural transformation. What the role means. What it requires. What it’s for.
Cognizant expected this inflection point in 2032.
It arrived in 2026.
The same week, the Snowflake CEO said something at his company’s annual summit that’s worth reading twice.
“This belief in the current set of job specializations and job descriptions is a complete myth. There are many jobs that are simply not needed, and there are new ones that will be needed.”
He wasn’t criticizing AI. He was describing speed.
Read the two data points together.
What “existential change” means for accountability
The discussion around AI disruption tends to run in one direction: which tasks survive automation?
That’s the wrong frame for 30% of the workforce.
When a role faces existential change — not task substitution, but role reconstitution — the question isn’t which tasks survive. The question is: who owns the things that were previously attached to that role?
There’s a gap between the role as-written and the role as-lived that was always present. But it was stable enough to manage. The accountant who absorbed informal analysis. The product manager who became the de facto support escalation path. The marketing director who quietly owned competitive intelligence because nobody else was looking.
None of that is in the job description. It lives in the person’s daily practice. In their inbox. In the things they said yes to over time.
When the role transforms — when AI absorbs the routine outputs and the human is left with a different, less defined surface of responsibility — those informal accountabilities don’t automatically go anywhere.
They either stay with the person, now under pressure to redefine what they actually do.
Or they fall into the gap.
The loop stays open. Nobody owns it.
The governance answer and why it misses the point
At the same COO Summit, Cisco’s chief people officer described how her team thinks about AI governance.
“You don’t just let any person into your home, talk to your children, eat your food, sleep in your bed. You ask them who they are and why they’re there. It’s the same thing with our work.”
That’s the right intuition for the wrong problem.
The governance frameworks emerging around AI answer one question: what is this system allowed to do? They create guardrails around AI behavior. They protect company data, compliance posture, risk exposure.
They don’t answer a different question: given what’s now being done by AI systems, what is each human being specifically accountable for?
That’s not a governance question. It’s a coordination question.
An org chart says who reports to whom. A job description says what the role is supposed to do. A governance framework says what the tools can do.
None of these systems tracks what each person actually owns — in real-time, as roles evolve faster than job descriptions can catch up.
What the Snowflake CEO is really naming
When Ramaswamy says job specializations are a myth, he’s making a structural claim about speed. The abstractions that organize human work are becoming unreliable faster than the institutions maintaining them can refresh.
But this observation has an uncomfortable corollary.
If job descriptions no longer reflect reality, then the system that tracks individual accountability has to be something else.
Not the org chart — it’s a reporting hierarchy, not a responsibility map.
Not the task management system — it tracks tasks, not accountability for outcomes.
Not the AI agent — it handles execution, not ownership of loops.
The missing layer is a system that knows what each person has actually taken on. What they’ve committed to. What they’ve implicitly inherited through the evolution of their role. What’s now open because the role changed faster than anyone was tracking.
The accountability gap compounds
Here’s what makes this harder than it looks.
When a role changes and an accountability goes untracked, it doesn’t show up immediately. The loop stays open quietly. Work proceeds. Other things get prioritized. Weeks later, someone notices the thing nobody finished. By then, it’s unclear who should have owned it.
In a stable org, you could trace it. The job description gave you a starting point. The manager conversation filled in the rest.
In an org where 30% of roles are in existential transition simultaneously — where “what your job is” is genuinely unclear and shifting week by week — that traceability breaks down.
The coordination overhead doesn’t decrease as roles evolve. It increases. More ambiguity means more checking, more clarifying, more status communication, more meetings to figure out who owns what. The work about work expands precisely when people have the least capacity to absorb it.
Cognizant’s researchers initially predicted this threshold in 2032. The people designing the coordination infrastructure to handle it are still operating on the 2032 timeline.
What a coordination layer would have to know
The layer that doesn’t exist yet isn’t a governance tool and it isn’t a task manager.
It’s a system that tracks the difference between the role as-described and the role as-lived. That knows what each person has implicitly taken on. That notices when a commitment is open and unowned. That routes accountability when roles shift — not after the incident, but before the loop goes dark.
Most organizations don’t have this. Most individuals don’t have this. The tracking is manual, informal, and leaky by design.
That’s not a people problem. Motivated, capable workers are doing the tracking in their heads.
It’s an infrastructure problem. There is no system above the individuals doing what a coordination layer should do.
Thirty percent of the workforce is in existential transition. Their open loops are going with them.
The loop doesn’t close itself.
And if nobody owns it, it doesn’t close at all.
