What Happens If AI Changes the Work of Saudi Arabia's Foreign Workforce?

Every time AI comes up in a majlis or on X, someone asks: "Is it going to take our jobs?"

I want to add another question: if automation changes the work done by foreign workers in Saudi Arabia, who captures the productivity benefit?

This is a scenario to examine, not a forecast that an entire workforce will be replaced. A job usually contains many tasks. Making one task faster does not necessarily remove the role, reduce headcount, or create a new Saudi job.

Remittances are context, not an automation savings budget

Outward personal transfers are a useful part of the economic picture. The Saudi Central Bank publishes transfer data in its Monthly Statistical Bulletin.

But a national remittance total is not a pot of money that a company can reclaim by buying AI. It spans workers, incomes, and sectors with very different automation possibilities. To assess one business, start with its actual tasks, labor costs, and service requirements.

Even if an employer spends less on a task, the saving does not automatically become domestic wealth. Software licences, imported equipment, cloud services, integration, energy, maintenance, supervision, and training all have costs. Some payments may go abroad. A locally hosted service can still rely on foreign hardware or suppliers.

The question is therefore conditional: after those costs, does the change create more useful work, better wages, stronger local capability, or a better service?

Sector by sector: test tasks before predicting jobs

These are candidate use cases, not claims that whole occupations are ready to disappear:

• Construction and contracting: compare planned and actual progress, organize inspection records, or flag missing documents. Physical work and safety responsibilities need separate assessment. A digital building model is not, by itself, an autonomous construction system.
• Retail and warehousing: draft product information, sort support requests, or flag inventory exceptions. Returns, unusual stock conditions, and customer disputes still need an accountable operator.
• Administration: extract fields from standard documents, prepare reconciliation suggestions, or draft routine correspondence. Check error rates and exception handling before expanding authority.
• Transport and logistics: compare routes, summarize delivery exceptions, or prepare dispatch recommendations. Autonomous driving is a separate technical and regulatory question, not an automatic next step from better routing.

In each case, the pilot should measure a complete workflow. A faster first draft can still leave the team with more correction, review, or follow-up.

Who benefits is a design decision

A productivity gain could fund lower prices, higher wages, better service, more hiring, or higher profits. The technology alone does not choose among them.

For founders, the practical decisions are:

1. Give a named person ownership of the workflow and its exceptions.
2. Train the team to inspect outputs, maintain approved data, and manage permissions.
3. Compare total operating cost with the current process, including review and support.
4. Decide how a verified gain will improve the business and the people doing the work.
5. Build transferable local capability instead of dependence on an opaque vendor.

Saudi workers can develop valuable skills in process design, quality review, customer relationships, data stewardship, and system operations. Those skills need training and real demand; a new job does not appear simply because an algorithm exists.

The opportunity I would pursue

My view is that Saudi businesses should compete on better systems and stronger skills, rather than on repeating routine work more cheaply.

Start with one painful task. Keep human responsibility visible. Measure the result after the full cost of the system. Then ask whether the benefit is reaching the people, customers, and local capabilities you intended to support.

That is a more useful ambition than treating the foreign workforce as one replaceable block or treating remittances as guaranteed savings.

For an illustrative economic design, see the Saudization mechanism page. Its calculator scenarios are proposals to examine, not proof of what automation will deliver.