In January 2026, the IMF updated its analysis of AI’s impact on the global labour market. The headline figure: 40% of jobs worldwide are exposed to AI, rising to 60% in advanced economies. But “exposed” doesn’t mean “eliminated.” The reality is more nuanced — and more actionable — than the headlines suggest.
À retenir
- 40% of global jobs are exposed to AI — but exposure means transformation, not always elimination
- Routine cognitive tasks face the highest displacement risk, not just manual labour
- The skills that protect workers are judgement, relationship management, and domain expertise combined with AI fluency
- Organisations that invest in upskilling see 2-3x better AI adoption outcomes
The data: which sectors face the greatest impact
The OECD’s 2025 Employment Outlook and Goldman Sachs’ updated AI economics research converge on a clear picture. AI’s impact is concentrated in white-collar, knowledge-intensive roles — not the blue-collar jobs that previous automation waves targeted.
High exposure (50%+ of tasks automatable):
- Administrative and office support
- Financial services (analysis, reporting, compliance checking)
- Legal services (research, contract review, due diligence)
- Customer service and call centres
- Content creation and basic copywriting
- Data entry and processing
Moderate exposure (25–50% of tasks automatable):
- Marketing and communications
- Software development (testing, documentation, basic coding)
- Human resources (screening, scheduling, policy drafting)
- Accounting and bookkeeping
- Translation and localisation
Lower exposure (under 25% of tasks automatable):
- Healthcare (clinical decision-making, patient care)
- Skilled trades (plumbing, electrical, construction)
- Teaching and education (in-person delivery)
- Social work and counselling
- Creative direction and strategy
60%
of jobs in advanced economies are exposed to AI — with roughly half of those likely to benefit from AI augmentation rather than displacement
Source : IMF Global AI Labour Impact Study 2026
Why this time is different
Previous automation waves — from the power loom to robotic assembly lines — primarily affected repetitive physical tasks. AI breaks this pattern because it targets cognitive tasks: reading, writing, analysing, summarising, predicting.
This means the workers most affected are often well-educated, well-paid professionals who never considered their jobs “automatable.” A junior lawyer spending 60% of their time on document review. A financial analyst who builds models from structured data. A marketing coordinator who writes first-draft copy. These tasks are exactly what large language models do well.
The biggest risk isn’t that AI replaces your job. It’s that someone who knows how to use AI replaces you. The World Economic Forum projects that 83% of employers plan to prioritise AI-skilled candidates by 2027.
The skills that protect you
If AI handles routine cognitive tasks, the value of human workers shifts to what AI cannot do — or cannot do reliably:
1. Judgement under uncertainty. AI can analyse data and suggest options. It cannot make high-stakes decisions that require weighing incomplete information, ethical considerations, and organisational context. Senior professionals who combine domain expertise with AI tools become more valuable, not less.
2. Relationship management. Negotiation, client relationships, team leadership, and stakeholder management remain fundamentally human skills. AI can prepare the brief; a human must read the room.
3. Creative direction. AI generates content at scale. But deciding what to create, why, and for whom — strategic and creative thinking — remains a human function. The art director, not the graphic designer, is the more defensible role.
4. AI fluency. The most powerful skill isn’t resisting AI — it’s mastering it. Workers who understand prompt engineering, know how to verify AI outputs, and can integrate AI into professional workflows become the most productive people in any team.
5. Domain expertise + AI. A mediocre lawyer using AI outperforms a great lawyer ignoring it. But a great lawyer using AI is in a category of one. Deep domain knowledge combined with AI fluency is the highest-value combination.
What the UK government data shows
The UK’s Department for Science, Innovation and Technology (DSIT) published its updated AI exposure analysis in late 2025. The findings for the UK specifically:
- Professional, scientific, and technical roles — highest exposure at 46% of tasks
- Financial and insurance services — 43% of tasks exposed
- Public administration — 38% of tasks exposed
- Education — 27% of tasks exposed (mostly administrative, not teaching)
- Healthcare — 22% of tasks exposed (mostly administrative and diagnostic support)
The UK data confirms a critical insight: exposure is highest in sectors that are also subject to AI regulation under the EU AI Act and emerging UK frameworks. This creates a dual imperative — manage the workforce transformation while meeting compliance requirements.
83%
of employers plan to prioritise AI-skilled candidates by 2027
Source : World Economic Forum Future of Jobs Report 2025
What organisations should do
The worst response is to do nothing and hope it sorts itself out. The second worst is to panic and start cutting headcount. The right approach is structured workforce transformation:
1. Assess your AI readiness. Before you can manage the impact of AI on your workforce, you need to understand your starting point. A structured AI readiness assessment evaluates skills, processes, data maturity, and governance across the organisation.
2. Map roles to AI impact. Not every role in your organisation faces the same level of disruption. Map each role against task-level AI exposure. This tells you where to invest in upskilling, where to redesign roles, and where natural attrition may be sufficient.
3. Invest in AI training — now. The organisations seeing the best outcomes from AI adoption are those that invested early in workforce AI training. This means practical, role-specific training that teaches employees to work effectively with AI tools, not theoretical courses about machine learning.
4. Build an AI competency framework. Define what AI competency looks like at every level of your organisation — from basic literacy to advanced integration. This gives employees a clear path for development and gives leaders a way to measure progress.
5. Meet your compliance obligations. The EU AI Act requires organisations to ensure AI literacy among their workforce. ISO 42001 provides a governance framework. Workforce transformation and compliance aren’t separate workstreams — they’re the same programme.
Start with a pilot. Choose one department with high AI exposure, run a structured training programme, measure the outcomes (productivity, quality, employee confidence), and use the data to build the business case for organisation-wide deployment.
How Brain helps
Brain prepares your entire workforce for the AI transition. Role-based training modules teach employees to use AI effectively, recognise risks like shadow AI, and work within your organisation’s governance framework — through practical exercises that take minutes, not days.
The result: a workforce that’s prepared for AI-driven change, measurable skill development across every team, and documented compliance with regulatory requirements.
Related articles
AI Change Management: 5-Phase Adoption Playbook
Overcome resistance, build communication strategies, phase training and measure AI adoption. Practical change management playbook.
AI Decision Making: 5 Ways Leaders Make Better Choices
Use AI for scenario planning, risk assessment, and bias mitigation. A practical guide to human-AI collaboration in executive decision making.
AI for CEOs: Executive Guide to Leading AI Adoption
Set the vision, build readiness, and govern AI across your organisation. A strategic AI adoption framework for chief executives.