AI skills by job function in marketing, finance, HR and operations

Indian enterprises expect to put generative AI into operations first, customer service second and marketing third over the next 12 months, so the AI skills a team needs differ by function. EY and CII's survey of 200 organisations, published on 16 November 2025, found 47% already run multiple GenAI use cases in production, and the top functions named for GenAI over the next 12 months are operations (63%), customer service (54%) and marketing (33%).

Is everyone going into AI and ML, or are business functions adopting it first?

Business functions use AI far more widely than job listings demand it. A PIB backgrounder of 12 February 2026, citing the World Bank, says 5.8% of India's white-collar job listings in 2025 required AI expertise, roughly 1 in 17. The same backgrounder cites the NASSCOM AI Adoption Index as finding 87% of enterprises in India actively use AI solutions. Use is far wider than the share of listings that require AI expertise.

Where the AI postings sit is also less concentrated than the headlines suggest. foundit's December 2025 tracker put IT software and services at 37% of AI postings in 2025, banking, financial services and insurance at 15.8% with the fastest growth at 41%, and manufacturing at 6% with 34% growth. That leaves most AI postings outside IT services.

The EY and CII survey adds a brake on enthusiasm. More than 95% of the surveyed organisations allocate less than 20% of their IT budgets to AI, and only 4% have crossed 20%. A non-technical manager will often be asked to deliver results from a small budget rather than to lead a large programme.

How is AI changing marketing work?

Marketing writes a lot, and writing is where the first measured gains appeared. MIT News reported in July 2023 on a Science study of 453 college-educated professionals, marketers among them, in which ChatGPT access cut the time on occupation-specific writing tasks by 40% and raised independent quality ratings by 18%. The tasks were 20 to 30 minutes long and online, so the figures describe drafts, not campaigns.

The management layer is visible in a job advert. A Databricks posting for a Director, Marketing Strategy and AI Transformation (United States, updated 6 October 2026) asks the post-holder to drive agent adoption across the global marketing organisation, set org-wide standards for prompts, context and agent workflows, and establish guardrails covering quality, risk, data privacy and responsible use. It asks for planning and governance, not for code.

Budgets show a gap between visibility and value. The MIT Project NANDA preliminary report (July 2025) cites an executive estimate that 50% of GenAI budgets go to sales and marketing, yet its best documented savings came from replacing BPO and external agency spend in back-office operations. The report rests on interviews at 52 organisations, so it is a signal and not a census. The existing list of AI tools for digital marketers covers the tooling side.

How is AI being used in finance, and where does autonomy stop?

The nearest evidence for finance teams is sector-level. Banking, financial services and insurance had the fastest-growing share of AI postings in foundit's 2025 data, at 41% growth, and Indeed Hiring Lab found AI mentioned in 18% of insurance postings in India in September 2025.

Seniority matters in this function. A Databricks posting for an AI Transformation Leader (United States, updated 21 September 2026) describes advising the non-technical C-suite and functional leaders such as the CFO and CMO on AI transformation roadmaps, and asks for at least 15 years of experience. IBM's 2026 CEO study found 64% of surveyed CEOs are comfortable making major strategic decisions based on AI-generated input, which raises the cost of unchecked numbers.

Autonomy has limits. The Harvard Business School experiment with 758 BCG consultants (2023) found users were 19% less likely to produce a correct solution on a task outside AI's frontier. Gartner predicts more than 40% of agentic AI projects will be cancelled by the end of 2027 because of costs, unclear value or weak risk controls, as reported by W.Media on 27 June 2025. That is a forecast.

A finance team's literacy therefore includes knowing which steps still need a named human sign-off.

What does AI mean for HR and people teams?

HR professionals were among the 453 participants in the writing-task study, so document-heavy work such as job descriptions and policy drafts falls within what it measured. The bigger shift is the training gap that people teams are expected to close. Naukri's AI at Work report of 16 July 2026 found only 35% of professionals say their employer actively provides AI training, and BCG's 2025 survey found only 36% of employees feel adequately trained.

Expectations are moving towards managers. In Microsoft's India Work Trend Index findings (20 August 2025), 63% of managers expect AI training to become a core team responsibility within five years. Microsoft's global 2025 Work Trend Index adds that 32% of managers are considering an AI Trainer role and 28% are considering AI workforce managers to lead teams of people and agents. Those are hiring intentions, not vacancies.

For people teams the literacy that matters is practical: knowing what guidance employees get, which AI use is authorised and how learning is tracked. The World Economic Forum's Future of Jobs Report 2025 lists talent management among the top 10 skills on the rise to 2030, a global employer view.

Where are the measured gains in operations and customer service?

Customer support has the best field evidence. Brynjolfsson, Li and Raymond studied 5,179 customer support agents and found a generative AI assistant raised issues resolved per hour by 14% on average and by 34% for novice and low-skilled workers, with minimal impact on experienced workers (NBER Working Paper 31161, revised November 2023). It covers one company's support staff, so it does not transfer automatically to managers or to other tasks.

Operations is where Indian leaders expect to deploy next. In Microsoft's India findings, 59% of surveyed leaders say they already use agents to automate workstreams across entire teams, and 93% intend to use agents to extend workforce capacity within 12 to 18 months. Observe.AI's AI Agent Engineer posting (Hyderabad, 23 September 2026) shows the delivery side: designing, integrating and deploying enterprise voice and chat agents, with evaluation frameworks, "not building foundational AI models". Whether such a role needs code is covered in do you need coding to learn AI.

The MIT NANDA preliminary report points the same way for savings: its best documented results came from back-office operations rather than front-office tools. The existing overview of generative AI use cases across industries gives more examples.

What does AI literacy look like at a desk in each function?

The table lines up the function-level evidence. The final row comes from foundit's sector data rather than from a function survey.

To find a first use case, a manager can pick the most repetitive document or review step in the function and measure it before and after. Take a hypothetical finance executive who reconciles vendor statements each month. The first use case is a draft summary of exceptions, with the reconciliation itself still signed off by a person.

For managers whose first use case turns into a built tool, how to test an AI tool you built covers evaluation. One structured route is the Universal AI Program by MIT Open Learning, a self-paced online programme from MIT Open Learning, delivered in India in cooperation with Masai on the MIT Learn platform, that builds AI fluency without needing a coding background. Its modules cover Python, machine learning, deep learning, large language models, generative AI and ethics, plus industry-specific modules.

Frequently asked questions

How is AI being used in finance?

The evidence is thinner than in support or marketing. Banking, financial services and insurance grew fastest in foundit's 2025 AI postings, at 41%, and AI appeared in 18% of insurance postings on Indeed in September 2025. Autonomy remains limited: Gartner predicts, as reported by W.Media, that more than 40% of agentic AI projects will be cancelled by the end of 2027.

What are some examples of AI literacy?

Knowing which tasks suit AI and which do not, checking output against a source, and keeping private data out of public tools. In the Harvard Business School experiment, consultants were 19% less likely to be correct on a task outside AI's frontier, which shows why the first example matters. A manager setting review rules for a team is the same literacy at a wider scale.

What are the top 5 non-technical skills for working with AI?

Analytical thinking, judgement about when to trust output, workflow design, communication with technical teams and responsible use. The World Economic Forum's January 2025 report puts analytical thinking first, with seven in 10 companies calling it essential. The others reflect what employers such as Databricks write into postings: standards, governance and adoption. No single published survey ranks all five.

Which skills matter most after a B.Com in finance or accounts?

The evidence points to domain depth plus AI literacy. IBM's 2026 CEO study found 85% of respondents say all functional leaders must become technology experts in their domain, and the World Economic Forum ranks analytical thinking as the top core skill. Neither source ties a skill to a salary. The fair reading is that the finance knowledge stays the base.

Which function will use AI most in the next year?

Among the 200 organisations EY and CII surveyed, operations leads at 63%, then customer service at 54% and marketing at 33%. These are stated priorities for the next 12 months, not deployments. HR and finance did not appear in that top group, though both face the same training gap, with only 35% of professionals reporting employer-provided AI training in Naukri's 2026 report.

Next step

The Universal AI Program, MIT Open Learning page sets out the curriculum and how to apply. Related reading: What enterprise AI adoption in India means for managers in 2026 and Masters in AI for non-technical students, or is AI literacy enough.