The debate around AI and jobs is often framed in extremes. Some predict massive job destruction. Others argue AI will unlock new productivity and employment.
A new international study from Snowflake and Omdia adds an interesting perspective to that debate.
It suggests that, at least inside many organizations, AI adoption is currently associated with more job creation than job loss. But the findings also reveal a more complex transformation of work.
What the Snowflake study found
The report, The ROI of Gen AI and Agents, surveyed 2,050 business and technology leaders across 10 countries about the impact of AI adoption in their organizations.
The results are striking.
77% of organizations report AI-driven job creation
46% report job reductions linked to AI
42% say AI has only created jobs
11% say AI has only eliminated roles
35% report a mix of both
Among companies experiencing both hiring and reductions, 69% say the net effect on employment has been positive.
These findings suggest that AI may be reshaping organizational structures rather than simply shrinking them.
Where jobs are growing
The study identifies the strongest job growth in technical and infrastructure-related roles.
The largest increases were reported in:
IT operations (56%)
Cybersecurity (46%)
Software development (38%)
These roles are expanding as companies deploy AI systems, manage data infrastructure, and strengthen security around AI-driven environments.
In other words, AI adoption often creates demand for new technical capabilities inside organizations.
Where AI is reducing roles
At the same time, the study also finds reductions in several functions.
The most affected areas include:
IT operations (40%)
Customer service and support (37%)
Data analytics (37%)
This dual effect is particularly interesting.
The same functions experiencing the most growth are also seeing the most disruption.
This pattern suggests that AI is transforming tasks within functions rather than eliminating entire professions.
The real bottleneck: data
One of the most revealing findings in the report is not about jobs at all.
It is about data readiness.
Organizations report major challenges in preparing data for AI:
65% struggle with data silos
62% with data quality
62% with preparing data for AI use
Even more striking: the report estimates that only around 7% of unstructured data is currently AI-ready.
This highlights a key constraint on AI adoption.
For many companies, the real barrier is not the technology itself but data infrastructure and governance.
AI adoption is still early
The report also indicates that companies are still early in their AI maturity.
Organizations that actively measure the impact of AI report an average return of $1.49 for every dollar invested in AI initiatives.
But at the same time, 96% of respondents say they still face major challenges scaling AI across the enterprise.
This suggests that the large-scale organizational impact of AI may still lie ahead.
Important limitations
While the findings are interesting, they should be interpreted carefully.
This study reflects self-reported experiences from business leaders, not direct measurements of labor market outcomes.
Other research points to more mixed or uncertain long-term effects.
For example a the International Monetary Fund (2026) suggests that around 40% of global jobs are exposed to AI-driven change, while labour-market data indicates that AI may both augment and replace tasks within jobs rather than eliminate entire occupations.
AI may simultaneously create jobs, transform roles, and reduce demand for certain tasks.
The deeper shift: task transformation
Perhaps the most important takeaway is this:
AI does not automate entire jobs. It automates tasks within jobs.
When tasks change, organizations must redesign:
roles
teams
workflows
skills
and talent management
This means the real challenge of AI is not simply employment levels.
It is how organizations reorganize work.
The real question
So the debate about AI and jobs may be asking the wrong question.
Instead of asking whether AI will destroy jobs, the better question may be:
How will organizations redesign work when AI can perform a growing share of tasks?
That is where the real transformation of the labor market will happen.
What do you see in your organization? Is AI already changing roles and teams, or is the impact still limited?
