■ SAFE RISK ■ Public Service & Government
No, though the caseload management around them is automating quickly. The job's substance is walking a bewildered family through a system that assumes knowledge they do not have.
“AI processes paperwork. But welcoming a family to a new country needs a human smile.”
Our AI replacement risk score — how we score jobs
Resettlement casework is logistics with a human face. In the first ninety days a worker meets a family at the airport, secures housing in a market that discriminates against them, enrols children in school, arranges health screenings and immunisations, applies for social security numbers and benefits, sets up English classes, teaches how buses and thermostats and rent work, and pushes toward employment because funding timelines are brutal. Then comes the longer tail: immigration paperwork toward permanent status, interpretation, mental health referrals for torture survivors, mediating with landlords and employers, and advocating when a system rejects someone for a reason nobody explained.
Automation is genuinely useful here. Machine translation has transformed initial communication across languages agencies cannot staff for. Case management platforms, document generation for benefit and immigration filings, appointment scheduling and eligibility screening are increasingly automated. Job-matching tools connect clients to employers. Chatbots answer routine orientation questions in a client's own language at any hour. Agencies also use analytics for placement matching to communities with suitable jobs and diaspora support.
The remaining work is trust and friction. Machine translation fails on trauma disclosure, legal nuance and cultural framing where getting it wrong has immigration consequences. Landlord negotiation, school advocacy, and sitting with a family whose asylum claim was denied are relational. Funding volatility, not automation, is what actually threatens these jobs — which is why our score sits at 6. A concrete example of the limit: a landlord who has decided he does not rent to families like this. No system resolves that. A caseworker with three years of local relationships calls a different landlord who owes her a favour, and a family has an apartment by Friday.
Automatability: our editorial assessment of current and near-term AI capability
Resilient for the foreseeable future, with real change in the mix of work. Translation, forms and case documentation are automating now, which should let caseworkers carry heavier loads through the late 2020s — and agencies under funding pressure will expect exactly that. The threat to employment in this sector is political: admission quotas and grant funding swing violently with governments. Displacement itself keeps rising.
For routine matters it already handles a great deal — appointment reminders, orientation materials, basic questions — and that is a genuine gain for agencies that cannot staff every language. For asylum interviews, medical and mental health conversations, and legal advice it is unsafe: mistranslation carries immigration and clinical consequences, and trauma disclosure depends on trust that a phone screen does not build.
Automation is compressing the paperwork half of the role, so agencies will expect higher caseloads per worker. But the sector's employment swings are driven by politics and funding — admission ceilings, grant cycles, government changes — far more than by technology. Global displacement continues to rise, so the underlying need grows. Expect the same or fewer workers doing more relational work.
In the first months: airport pickup, housing setup, school enrolment, health screenings, benefits applications, English class registration, bus and budgeting orientation, and job placement against a hard funding clock. Afterwards: immigration filings toward permanent status, mental health referrals, landlord and employer mediation, and advocacy when systems fail. It is part social work, part logistics, part translation, part fighting bureaucracy.
Trauma-informed practice, immigration process knowledge, and housing and employment advocacy — the parts requiring negotiation with real institutions. Add fluency with case management systems and document automation, since agencies are consolidating on those tools. Language skills remain valuable but their competitive edge is eroding for routine use; cultural brokerage, which is different from translation, is not.