AI and the future of Business and Enterprise work: what changes, what stays human

Key facts

• Clerical occupations are the most exposed group in the ILO's 2025 index. Accounting and bookkeeping clerks score 0.64 out of 1, accountants 0.51 and managing directors 0.38. Global generative-AI task-exposure estimates; not a Malaysian forecast; not a job-security rating^[1].
• In Malaysia, the Human Resources Minister said in August 2025 that about 245,000 wholesale and retail jobs and 89,000 global business services jobs are expected to be significantly impacted within three to five years, based on a TalentCorp study. "Impacted" does not mean lost^[2].
• Employers in Malaysia in the World Economic Forum (WEF) survey expect data entry clerks and administrative assistants to shrink by 2030, and AI specialists and supply chain and logistics specialists to grow. Expectation of employers in Malaysia in the WEF's global survey, published 2025^[3].
• 77% of WEF employers plan to retrain staff to work alongside AI, while 41% expect to shrink their workforce as AI does more. Global employer expectation, published 2025^[3].

Business runs on numbers, documents and conversations. In the ILO's index, the routine office occupations score highest and the selling and leading ones lower. Malaysian figures here are projections, so treat this page as questions to test with people who run businesses.

For routes in, SPM subjects and professional qualifications, start with Business & Enterprise after SPM.

How to read this page

Observed means someone measured it. Projected means someone estimated the future, so we say who, when and for where. Projections are often wrong.

The scores come from the International Labour Organization (ILO). They estimate how much of an occupation's current tasks generative AI (tools that write, summarise or make images from a prompt) could potentially do, from 0 to 1^[1]. "Not exposed", the lowest band, is a technical label: most of the occupation's tasks scored consistently low for what generative AI can do today^[1]. It does not mean AI will never change the work. Exposure is not job loss either: the ILO says that depends on adoption and on whether workers get the chance to learn the tools^[1].

What is already happening (observed)

• Customer service, in US data. In Anthropic's March 2026 data, customer service representatives had the second-highest observed AI coverage of any job, after computer programmers, and data entry keyers were at 67%^[4]. US usage data, not Malaysian demand.
• Marketing content is mostly work. In June 2026, Anthropic reported that 80% of conversations creating marketing content were work-related^[5]. Anthropic usage data from a sample it says is not representative.
• Malaysia. We did not find a current Malaysia-specific study that measures AI use across Malaysian businesses. The Malaysian figures below are projections.

What people project, and who says so

• TalentCorp, for Malaysia (reported November 2024 and August 2025). TalentCorp, the Ministry of Human Resources' talent agency, studied 10 sectors in 2024. Of about 3.5 million jobs, it rated 18% (620,000) as highly impacted, meaning they may merge with other roles or be displaced, and 53% as still relevant but needing upskilling^[6]. The minister later said the largest numbers were in wholesale and retail trade (245,000), food manufacturing services (203,000) and global business services (89,000), the shared-service centres that handle finance, HR and customer support for companies^[2]. We read news reports of the study, not the report itself.
• Employers in Malaysia. They put AI and machine learning specialists, supply chain and logistics specialists, factory workers and managing directors among the roles they expect to grow, and administrative assistants and data entry clerks among those they expect to shrink. They said people mainly do 47% of work tasks today and expected 33% by 2030. Expectation of employers in Malaysia in the WEF's global survey, published 2025^[3].
• Professional services worldwide. Employers said people mainly do 52% of tasks today and expected 31% by 2030. Business intelligence analysts and project managers were growing roles; accountants and auditors, and bookkeeping clerks, were shrinking. Global employer expectation, published 2025^[3].
• Shrink or retrain. 77% plan to retrain staff to work alongside AI by 2030, and 41% expect to shrink their workforce. The two add up to more than 100%, so many employers expect to do both. Global employer expectation, published 2025^[3].

Task by task

These cards are our reading of the sources for a classroom exercise, not a forecast. A practitioner may disagree, and their reasons are worth hearing.

Record transactions and reconcile accounts

• What a person does: deals with the odd entries and applies the rules.
• What AI might assist: matching most entries automatically.
• What this source does not show: how many bookkeeping jobs Malaysian firms will cut or change.

Prepare financial reports and advise a client

• What a person does: signs off, explains and takes responsibility.
• What AI might assist: drafting reports and spotting patterns.
• What this source does not show: how professional bodies will treat AI-made work.

Answer customer questions by chat or phone

• What a person does: takes complex cases and upset customers.
• What AI might assist: routine questions through chatbots.
• What this source does not show: how Malaysian shared-service centres will staff this.

Sell face to face and negotiate a deal

• What a person does: reads the other person and builds trust.
• What AI might assist: preparing figures and talking points.
• What this source does not show: anything about trust, which no score measures.

Write marketing posts and product descriptions

• What a person does: checks claims are true and fit the brand.
• What AI might assist: fast first drafts.
• What this source does not show: what small Malaysian firms will pay for this work.

Price an offer for a small business

• What a person does: checks costs, margins and what customers will actually pay.
• What AI might assist: drafting the calculation.
• What this source does not show: how often AI gets local costs wrong.

New tasks that are appearing

• Checking AI-produced figures, reports and contracts before they reach a client.
• Setting up and supervising AI customer-service tools, and handling what they cannot.
• Deciding which tasks to automate, and what happens to the staff who did them.
• Keeping customer data private. Malaysia's voluntary National Guidelines on AI Governance and Ethics (MOSTI, September 2024) include fairness, privacy and accountability among their seven principles^[7].

This list is our reading of the sources above, not an official one.

What stays human: hypotheses to test

These are practitioner hypotheses to test, not findings. The sources do not show that any of these tasks will stay unchanged.

• Does persuading a customer and keeping the relationship still need a person?
• Who makes a decision with incomplete information, and owns it?
• Who leads the team when things go wrong?
• Who spots the business opportunity nobody has written down yet?

Ask a business owner which of these hold for them, and which AI has already changed.

What this means for you

• Use clerical roles as a stepping stone, not a destination. Admin and data entry work is where the ILO's global scores and Malaysian employers' expectations both point to the biggest change.
• Get good with numbers, then check the machine. Knowing when a number is wrong is the skill to practise.
• Practise selling and presenting. Commercial thinking, presenting and negotiation are the skills listed for this area, and they involve people.
• Get real customer experience. A part-time job, a bazaar stall or a small online shop teaches you what customers want faster than a course.
• Watch TalentCorp's next sector studies, how Malaysian banks and shared-service centres use AI, and the WEF's next Future of Jobs survey.

Your next step

1. On Day 2, choose the Business and Enterprise taster (an offer and a price for a small business). Price it once with an AI assistant and once by hand, and note every number you had to correct. If your cohort runs four labs, the costing and pitching sections of the People and Society and Design and Create lab briefs cover the same skills.
2. Ask the practitioner: "Which tasks in your team has AI taken over, and what did the people who did them move on to?"

About the evidence

ILO scores (a global generative-AI task-exposure estimate; not a Malaysian forecast; not a job-security rating), on a 0 to 1 scale with the ILO's band^[1]:

• Accounting and bookkeeping clerks 0.64, gradient 4
• Financial analysts 0.62, gradient 4
• Contact centre information clerks 0.58, gradient 3
• Advertising and marketing professionals 0.55, gradient 3
• Accountants 0.51, gradient 3
• Commercial sales representatives 0.49, gradient 2
• Management and organisation analysts 0.46, gradient 2
• Shop sales assistants 0.38, gradient 1
• Managing directors and chief executives 0.38, minimal exposure
• Supply and distribution managers 0.38, minimal exposure
• Stock clerks 0.37, minimal exposure

Most of the 13 occupations in the top band, gradient 4, are clerical^[1]. ILO experts also rated one structured money task, developing a household budget, as highly automatable (0.64)^[1]. The bands run from "not exposed" and "minimal exposure" up to gradients 1 to 4. The ILO built its scores from a survey of 1,640 workers in Poland, rounds of discussion with international experts, and an AI assistant built on that input. Every task counts equally, and the ILO says local factors its global study does not cover can speed up or limit the change^[1]. ILO occupation groups do not map one-to-one onto Malaysian job titles, and we do not claim a direct mapping.

WEF figures are expectations from a survey of more than 1,000 employers across 55 economies, published January 2025^[3]. Anthropic figures are US or usage data, not Malaysian demand. The method behind the task cards is in Tasks, not job titles.

For facilitators

Use the task cards as a starting list for the Day 2 AI task X-ray and ask the practitioner to correct them before learners use them. Each learner supports one card with a source and challenges one assumption. Stop any conclusion that a job is "safe" or "dead": the evidence here does not support either.