The Workforce Brief · Jobs & Talent · 3 November 2026
An organisation can buy more AI capacity while running short of the people needed to make it useful. The pressure may appear in a queue of technical approvals, an overloaded supervisor or a project waiting for an experienced reviewer. For employers planning their next AI investment, specialist time deserves its own budget.
The new evidence: shortages persist alongside AI
The OECD’s Mapping Skill Shortages in OECD Countries, published on 14 October 2026, identifies persistent shortages among ICT, health and teaching professionals. Its updated database generally extends through 2024: this is newly published analysis of earlier labour-market evidence, not a live November vacancy count. The report links ICT demand to digital transformation and AI, without establishing that AI alone caused the shortages.
For Malaysian employers, these international findings are a prompt to investigate local constraints, not a Malaysian shortage estimate. Check accepted offers, recruitment lead times, overtime and project delays before drawing conclusions about your own workforce. A hard-to-fill vacancy can reflect scarce expertise, but also an unrealistic job specification or unattractive working conditions.
Find the bottleneck before adding another tool
Our practical interpretation is that AI investment should include a specialist-capacity review. Follow one piece of work from request to approval. Identify where it waits, who can release it and how much of that person’s time is consumed by preparation rather than judgement. Automating preparation helps only if the downstream reviewer can absorb the additional output.
Consider an illustrative AI data-centre project. An assistant may organise equipment documentation quickly, while a small engineering team still needs to resolve interfaces and approve changes. Producing twice as many documents does not create twice as much engineering capacity. A useful first intervention could be standardised submissions, fewer incomplete requests and scheduled review windows.
For AI infrastructure and talent planning, distinguish design expertise, commissioning responsibilities and operating coverage. Do not assume one impressive technical hire can cover every phase. This is an operating recommendation, not a claim about any Event Systematic project or a forecast of staffing demand.
Budget for review capacity, not just headcount
Create a weekly capacity sheet for each critical specialist. Record available hours after routine duties, time committed to reviews, mentoring commitments and backup cover. Compare those hours with the work arriving. Treat the resulting figures as planning assumptions to test, rather than universal utilisation targets.
Three measures make the sheet useful: time waiting for specialist review; first-pass acceptance of submitted work; and hours spent correcting avoidable defects. Read them together. Shorter queues achieved by skipping checks are not an improvement. Equally, a higher volume of AI-generated output can conceal a growing correction burden.
Set a trigger for action before the queue becomes a delivery crisis. Depending on the cause, the response might be a narrower intake, clearer submission standards, temporary specialist support or recruitment. Protect time for a second employee to learn the review process under supervision; otherwise today’s throughput target can deepen tomorrow’s dependency on one person.
Build skills around the scarce decision
The OECD’s AI and skills: What we know so far, dated 5 June 2026, describes skills shortages as an adoption barrier and highlights digital, data and complementary human capabilities. Its evidence supports broader workforce development; it does not promise that a short course can replace specialist experience.
An employer can translate that distinction into two learning tracks. The first helps colleagues prepare reliable inputs: checking sources, recording assumptions and recognising missing information. The second develops future reviewers through supervised decisions, explanations and feedback. Connect both tracks to the AI Academy discussion, with scope and assessment agreed for the actual work.
For talent, build a portfolio that shows how you reduced uncertainty for another professional. A well-documented handover, an identified discrepancy or a clear escalation can demonstrate value beyond a polished AI output. Use synthetic or authorised examples, and explain which decisions remained with your supervisor. Explore career development pathways around these capabilities.
Watch: skills for changing work
OECD’s official discussion, What skills are needed for the AI and green jobs era? (OECD channel, 6 July 2025), provides background on adaptation and upskilling. It is an expert discussion, not evidence of a particular employer’s results.
Turn the diagnosis into a staffing brief
Bring one bottleneck, its current workload and the decisions it delays to your next workforce review. Event Systematic’s employer services provide a starting point for discussing recruitment and capability needs. Contact us to scope your workforce priorities.
Explore The Workforce Brief. Sources checked 3 November 2026.
