Snowflake Research Finds AI-Driven Job Creation Outpaces Job Loss Globally

AI is creating more jobs than it is eliminating, according to a new global study from Snowflake and Omdia by Informa TechTarget – but the picture is far from simple, with Singapore in particular struggling to translate AI adoption into measurable business outcomes.

The research, titled “The ROI of Gen AI and Agents”, surveyed 2,050 business and technology leaders across 10 countries and found that 77 per cent of organisations report AI-driven job creation, compared to 46 per cent reporting job losses.

Among organisations experiencing both hiring and cuts, 69 per cent report the net impact of AI on their workforce has been positive.

AI Returns Are Real, But Challenges Persist

On a global level, the report paints a picture of AI delivering tangible value. Organisations report earning roughly $1.49 for every dollar invested in AI, and 92 per cent of early AI adopters report positive return on investment.

Businesses plan to allocate 22 per cent of their technology budgets to AI in the coming year, a signal that investment is accelerating rather than pulling back.

Yet the path to scale remains bumpy. A full 96 per cent of organisations surveyed report facing significant challenges in scaling their AI initiatives, with data quality, employee skills and integration with legacy systems emerging as persistent barriers.

Technical Roles See The Strongest Job Growth

The study reveals that AI’s workforce impact is concentrated in technical functions, where both the strongest job gains and steepest reductions are occurring.

IT operations leads the way, with 56 per cent of respondents reporting job gains in that area. Cybersecurity follows at 46 per cent, and software development at 38 per cent.

But IT operations is also where the most job losses are taking place, with 40 per cent of respondents reporting reductions. Customer service and support (37 per cent) and data analytics (37 per cent) round out the three functions most affected by AI-driven cuts.

The data suggests AI is reshaping these functions rather than uniformly expanding or shrinking them. As productivity increases, organisations are restructuring teams – automating certain tasks while adding new capabilities elsewhere.

“AI’s impact won’t be uniform – some roles will dramatically amplify their influence and productivity, while others risk being left behind,” Anahita Tafvizi, Chief Data Analytics Officer at Snowflake, explained. “The difference comes down to how effectively it’s used: breaking down problems with first-principles thinking and guiding AI agents like high-performing teams.”

Tafvizi added that the strongest ROI is coming not from experimentation alone but from embedding AI into core operations while strengthening data readiness and governance policies.

Singapore Keeps Pace On Adoption But Falls Behind On Outcomes

Singapore matches the global average in terms of AI activity, with 39 per cent of local respondents reporting their organisation has many generative AI use cases in place today.

However, the data reveals a gap when it comes to actual deployment within departments. Across almost every major business function, Singaporean organisations are applying AI less widely than their global counterparts.

In IT operations, 45 per cent of Singaporean organisations report AI adoption compared to 63 per cent globally. In software development, the split is 39 per cent versus 50 per cent. Cybersecurity stands at 33 per cent locally against 55 per cent globally.

The narrower deployment is correlating with more muted outcomes. Fewer Singaporean respondents report gains in operational efficiency (79 per cent versus 89 per cent globally) and cost reduction (75 per cent versus 82 per cent globally).

Only 33 per cent of Singapore respondents report quantified ROI from AI, well below the 49 per cent global average.

Budget allocations reflect a cautious approach. Singaporean respondents estimate generative AI will account for 15 per cent of their tech budget over the next 12 months – the lowest figure of any market surveyed, trailing the 23 per cent global average and leading markets like India (28 per cent), the UK (23 per cent) and the US (23 per cent).

Identifying the right use cases also appears to be a particular pain point locally, with 32 per cent of Singaporean respondents citing it as a top challenge, compared to just 19 per cent globally.

Local Leaders Point To Data Foundations As The Missing Piece

Jenny Koh, Country Manager for Singapore at Snowflake, described the local challenge as one of execution rather than ambition.

“Singapore’s reputation for technology adoption is well-earned, and with the National AI Strategy 2.0 shifting the national focus from adoption to impact, the stakes have never been clearer,” Koh observed. “Yet, with less than a third of local organisations reporting quantified ROI against half globally, it is clear that deploying AI and generating results are two different things.”

Koh pointed to the need for stronger data foundations, noting that in her conversations with local leaders, the recurring challenge is not a lack of vision but the infrastructure to support it.

“The organisations that will lead are those building the right data foundations deliberately, bringing AI to their data to solve specific business missions rather than pursuing it broadly,” she added.

Data Readiness Remains The Primary Bottleneck

Across all markets, the report identifies data readiness and governance as the primary constraints on AI scaling – not the technology itself.

Nearly eight in 10 respondents report experiencing technical or data-related challenges. Specifically, 65 per cent of respondents find it challenging to break down AI data silos, while 62 per cent report difficulties in measuring and monitoring AI data quality. Another 62 per cent struggle to prepare data to be AI-ready.

Only 7 per cent of respondents globally report that more than half of their unstructured data is AI-ready. India leads at 14 per cent, followed by Australia and New Zealand at 12 per cent and Canada at 10 per cent. The United States comes in at 8 per cent.

Governance is an equally pressing concern. The report found that 57 per cent of employees, including 66 per cent of C-level leaders, report using non-approved AI tools. Meanwhile, 60 per cent of respondents indicate their organisations need greater investment in data infrastructure and monitoring software.

AI Is Already Embedded In Core Functions

The research shows AI is no longer confined to experimentation. It is already active across major enterprise functions.

Some 62 per cent of IT operations teams report active AI use, along with 59 per cent of data analytics teams, 53 per cent of cybersecurity teams and 50 per cent of software development teams.

By contrast, functions such as procurement, sales and marketing are the slowest to adopt, with around 30 per cent of each reporting active use.

At the industry level, advertising and media leads with 42 per cent of organisations reporting AI in production, followed by healthcare and life sciences at 34 per cent, and both manufacturing and technology at 32 per cent.

Nearly half of all code – approximately 48 per cent – is now reported to be AI-generated. Organisations are reporting measurable benefits from AI coding tools, with 82 per cent citing improvements in code testing, bug detection and resolution, and 80 per cent reporting gains in overall code quality.

Maturity Drives Better Outcomes

The report underscores a clear link between AI maturity and workforce outcomes. Among organisations deploying AI across many use cases, 75 per cent report a net positive impact on jobs, compared to 56 per cent of those with more limited, early-stage deployments.

Adam DeMattia, Senior Director of Research at Omdia by Informa TechTarget, highlighted the importance of getting the foundations right.

“The data shows that AI is delivering tangible returns, but scaling it successfully requires a strong data foundation and governance framework,” DeMattia remarked. “Organisations that can unify their data, improve quality and operationalise AI responsibly will be best positioned to sustain ROI and workforce gains.”

Methodology

The report is based on responses from 2,050 respondents across 10 countries: Australia/New Zealand, Canada, France, Germany, India, Japan, Singapore, the United Kingdom and the United States. All respondents are influential in their organisation’s current and future AI purchases. The survey was conducted between August 13, 2025 and September 17, 2025.

The full report is available on Snowflake’s website.

Last Updated on March 16, 2026 by Nick Ross