Portugal AI Job Risk 2026: Which of 5.2 Million Workers Are Most Exposed?

5.2 million workers, Eurostat + INE (Instituto Nacional de Estatistica) 2025 data

Employment data comes from Eurostat lfsa_egai2d and INE - Instituto Nacional de Estatistica (Statistics Portugal), using ISCO-08 major group classifications. Data year: 2025, covering approximately 5.2 million workers (5,245.1K). INE conducts the Inquerito ao Emprego (Employment Survey) quarterly, providing ISCO-08 occupational breakdowns consistent with Eurostat standards. Portugal's economy has undergone substantial structural change since the 2010-2014 austerity period: tech, tourism, and financial services grew rapidly while traditional manufacturing contracted.

Portugal's informal employment rate sits around 20%, higher than the Western European average and somewhat closer to Southern European comparators like Italy and Greece. This matters for AI risk because informal workers are largely absent from official occupation counts - the 5.2M figure captures the formal workforce, and informal employment tends to concentrate in lower-AI-risk sectors (construction, agriculture, domestic services). The AI exposure figures therefore slightly overstate the risk for the formal workforce relative to the full working population.

The most AI-exposed occupations in Portugal

General and keyboard clerks score 9.0/10 - the highest of any occupation group in Portugal - covering 272,500 workers. These workers perform data entry, administrative document processing, scheduling, office correspondence, and accounting support tasks. In Portugal's banking sector (Banco Comercial Portugues, Novo Banco, Caixa Geral de Depositos) and in the growing Lisbon-based European headquarters of international companies, these roles are directly in the path of AI document automation. Portuguese wages remain lower than Western European averages, but not low enough to insulate against AI tools priced in dollars and euros.

ICT professionals score 8.5/10 AI exposure across 167,400 workers. Lisbon's transformation into a European tech hub accelerated after Web Summit relocated there in 2016, and the Golden Visa program (until its 2023 reform) attracted technology entrepreneurs and remote workers. Engineering centres for companies including Mercedes-Benz, Volkswagen Digital Solutions, Natixis, and hundreds of software startups now operate in Lisbon and Porto. AI coding tools - GitHub Copilot, Cursor, and similar - directly reduce the coding hours per feature that justify junior developer headcount. This risk is not speculative; it is already affecting hiring decisions in 2025.

Customer services clerks score 8.5/10 across 106,500 workers. AI chatbots and voice agents are deployed in Portuguese banking, telecoms (NOS, MEO), and e-commerce at increasing scale. Numerical and recording clerks score 8.5/10 across 93,900 workers - stock control, production recording, and transport booking roles all face RPA (robotic process automation) and AI competition. Business and admin professionals score 8.0/10 across 219,600 workers, and business associate professionals score 7.5/10 across 295,300 workers - Portugal's largest high-risk group.

Why Lisbon's tech cluster amplifies Portugal's AI exposure

Most analyses of Portugal's AI risk focus on the tourism sector - understandably, since tourism represents roughly 15% of GDP and employs hundreds of thousands of workers. But the more immediate AI risk concentration is in Lisbon's white-collar sectors. When Portugal had a financial crisis in 2010-2014 and required an IMF bailout, the structural response was to attract tech investment and high-value service companies. That strategy succeeded - but it also created a concentration of ICT and professional services workers who are now in the highest AI-exposure groups.

The Golden Visa program's tech hub provision attracted non-EU tech entrepreneurs, and Web Summit's presence turned Lisbon into an annual gathering point for European and global tech investors. This built a local startup ecosystem and attracted multinational tech operations. The engineers, product managers, and business development workers hired into these roles are now in the 8.0-8.5/10 AI exposure range. The same economic success that pulled Portugal out of its fiscal crisis created an AI vulnerability in its fastest-growing sector.

Compare this to Spain, where Barcelona and Madrid have a similar pattern of tech hub growth overlaid on a tourist economy. Or France, where Paris's enormous professional services sector drives a high weighted average. Or Netherlands, where Amsterdam's financial and tech concentration creates similar dynamics at a higher wage level. The pattern repeats across Western Europe: the cities that successfully modernised their economies in the 2010s are now exposed to AI in the 2020s.

The safest jobs from AI in Portugal

Cleaners and helpers score 1.5/10 AI exposure, covering 212,100 workers. The work is entirely physical, on-site, and requires constant situational adaptation in hospitals, offices, hotels, and homes across Portugal. No AI system replaces a cleaner in a Lisbon apartment block or a Algarve resort. Personal care workers score 2.0/10 across 247,600 workers - and this is Portugal's most structurally important buffer. Portugal's population is aging rapidly (median age approximately 46 years, one of the highest in Europe), and the demand for personal care - elder care, home care, residential care facility workers - is structurally increasing regardless of AI development. These roles are location-specific, relationship-intensive, and physically demanding in ways that current AI and robotics cannot replicate.

Drivers score 2.5/10 AI exposure across 208,900 workers. Portugal's geography - with significant rural areas, Atlantic coastal routes, and Algarve tourism transport - means driver employment remains substantial. The robotics risk for drivers is 7.5/10, but autonomous vehicle deployment in Portugal's road network (particularly in rural areas) remains distant. Metal and machinery trades workers score 3.0/10 AI exposure across 176,900 workers, with a robotics risk of 6.5/10 - safe from AI language model competition, but facing gradual automation in manufacturing contexts.

What this means for Portuguese workers right now

Portugal's risk velocity score is 8.0/10 ("Disruption accelerating - 2 to 4 years"). This is slightly lower than the 10.0/10 assigned to the UK, Germany, and the US - not because AI arrives more slowly in Portugal, but because Portugal's lower wage levels and the composition of its workforce include more jobs in sectors where the economic case for immediate AI replacement is not as compelling as in higher-wage markets. This is a relative buffer, not a protection.

Portugal's recovery resilience score is 6.5/10. The country has a solid university system (University of Lisbon, University of Porto, Nova SBE) and government programs encouraging tech sector growth, but labour market flexibility is more constrained than in Nordic countries, and retraining infrastructure is less developed than in Germany or the Netherlands. Workers displaced from clerical and administrative roles face a harder transition path than their counterparts in countries with stronger active labour market policies.

For the 272,500 general clerks and 219,600 business and admin professionals, the practical near-term risk is wage compression before outright displacement. Employers facing AI tools that can handle 60-80% of routine administrative tasks will hire fewer replacement clerks when vacancies arise, reduce overtime, and compress salaries - rather than immediately making large-scale redundancies. This wage compression is already detectable in Portuguese administrative sector wage data from 2023-2025. The displacement follows wage compression by 12-24 months in comparable European markets.

For a global comparison of where Portugal fits, see our US vs World AI job risk analysis, or compare directly with UK and US data to understand the wage-risk interaction across different markets.