Demand for AI skills has spread past engineering: at 6 of the 25 biggest employers we tracked, more than half of non-engineering job postings mention AI, from 89.5% at Anthropic to none at Gopuff.

Why are so many companies suddenly requiring AI skills for non-tech jobs?

The spread of AI into non-engineering job postings reflects a broader shift in how companies are redesigning work, not just hiring specialists. KPMG's survey of 400 large US company CEOs found that 86% are redesigning roles, career paths and team structures for human-AI collaboration, up from 65% a year earlier, and 79% are hiring people with AI and tech skills, up from 64%.⁴

Lightcast senior economist Elena Magrini notes that the biggest shifts in AI job posting requirements since 2025 have been in non-technical skills: demand for cross-functional collaboration nearly doubled between Q2 and Q3 of 2025, with similar surges in stakeholder management and operational efficiency.³ That pattern matches what the job postings here show: AI fluency is turning into a written requirement across functions, with 162 of the 261 employers asking for it in at least one posting. The chart's sector figures reinforce the point, with AI mentioned in 54.2% of enterprise software postings and 52.3% in cybersecurity, well above the 7.3% share Indeed Hiring Lab recorded across all US postings on Sep 30, 2026.

What is AI fluency and why do employers want it?

AI fluency is the ability to work effectively with AI-supported tools and information as part of a role, not a technical certification or a checklist of software skills. According to Indeed, it covers how candidates interpret AI-generated insights, recognise their limits, and apply judgment when deciding what to act on.² A marketing manager reviewing AI-supported campaign analysis, an HR partner using AI-summarised employee feedback, or an operations leader reading AI-generated forecasts all need this capacity, even though none of those roles is technical.²

Deloitte describes AI fluency as how skilled a person is "in understanding, using and effectively collaborating with AI tools," and argues it needs to become part of an organisation's culture rather than a one-off training exercise.⁵ Indeed frames it as distinct from experience with any particular tool: the focus is on reasoning, problem framing and adaptability as AI workflows change.² That is why employers such as GitLab, SentinelOne and Glean now state AI fluency, literacy or proficiency as a requirement in at least 90% of their postings, treating it as a baseline expectation rather than a specialist add-on.

Which workers and industries are most at risk from the AI skills shift?

Workers in roles with little AI exposure in their current job postings face the most immediate pressure to adapt. The chart shows e-commerce and retail postings mention AI in only 12.1% of cases, health in 23.9% and hardware, energy and transport in 21.1%, suggesting those sectors have not yet embedded AI fluency as a written requirement. Gopuff, Lyra Health and One Medical, whose openings are mostly store, warehouse and clinical jobs, show near-zero AI mentions.

The World Economic Forum's Future of Jobs Report 2025, cited by Iternal, estimates that if the global workforce were 100 workers, 11 would need training but are unlikely to receive it, putting roughly 120 million people at medium-term risk of redundancy.⁶ Lightcast's Magrini notes that employers are increasingly seeking professionals who can bridge teams, operationalise AI systems and build trust in their outcomes, skills that go beyond any single technical role.³ Workers in sectors where AI job postings remain rare may have less time to build those capabilities before employers begin to require them.

What do CEOs and analysts say about hiring for AI capabilities?

Walsh said the mix of people is what is changing, not the total count. At KPMG, he said, the firm needs more engineers and more people with AI skills, and is training its own 50,000 US staff rather than trying to buy all that talent externally.¹ On the pace of change, he said: "Looking three months out is kind of the benchmark now."¹ Lightcast's Magrini adds that the rapid rise of stakeholder, operational and trust-related skills in AI job postings offers "an early signal of how AI roles, and the workforce more broadly, are evolving in real time."³

Will AI skill requirements keep spreading to lower-wage and frontline roles?

Whether AI skill requirements spread to lower-wage and frontline roles is not yet visible in this data. The chart covers 261 mostly US and European tech employers; large employers with their own career sites, including Amazon, Google and Walmart, are not in it, and e-commerce and retail postings here mention AI in only 12.1% of cases. The employers where AI fluency is already near-universal, such as Anthropic at 89.5% and Databricks at 77.2%, are AI-native companies whose entire product is AI.

KPMG's survey found that 75% of CEOs have rethought the skills entry-level roles need because of AI, which suggests pressure could move down the seniority ladder even if it has not yet reached frontline job ads.⁴ Deloitte argues that AI fluency needs to become part of an organisation's culture at every level, and that employers should encourage its use and reduce the misconception that AI threatens a person's job.⁵ Whether that translates into explicit requirements in warehouse, retail or clinical postings remains a possibility rather than a documented trend in the current data.