Will AI Steal My Job? · Role analysis

Administrative
Assistant

O*NET 43-6014.00 ESCO: Administrative assistants
High exposure

Administrative assistants keep offices running — managing correspondence, scheduling meetings, maintaining files, preparing documents, handling travel arrangements, and supporting managers and teams with the organisational work that enables everything else. They are the operational backbone of most organisations, handling a high volume of routine coordination tasks.

Task Map

TaskAI impactWhy
Draft emails, letters, and routine correspondence 🔴 High exposure AI writing tools draft professional correspondence rapidly from brief prompts. This was a core, time-consuming task for administrative assistants and is now heavily AI-accelerated for anyone with access to AI tools.
Schedule meetings and manage calendars 🔴 High exposure AI scheduling tools (Copilot, scheduling assistants) and automated calendar management handle meeting coordination with minimal human involvement. This highly routine task is substantially automating.
Prepare reports, presentations, and documents 🔴 High exposure AI tools generate first-draft reports and presentations from data and briefs significantly faster than manual preparation. Document production that once took hours is now accelerated dramatically.
Maintain filing systems and records management 🔴 High exposure Digital document management systems with AI-powered tagging and retrieval largely automate file organisation. Manual filing and record maintenance is declining as a distinct task.
Handle travel booking and expense management 🔴 High exposure Travel booking platforms and expense management tools (Concur, TravelPerk) have automated most of this work. AI assistants are further reducing the human involvement required for routine booking tasks.
Greet visitors and manage physical office presence 🟡 Changing The physical welcome — greeting clients, managing deliveries, ensuring the office runs smoothly day-to-day — requires human presence that digital tools cannot provide.
Support senior staff with research and information gathering 🔴 High exposure AI research tools can gather, summarise, and present information from multiple sources faster than human research. Tasks like compiling briefing notes or researching background information are heavily AI-assisted.
Coordinate across teams and manage office logistics 🟡 Changing Complex multi-party coordination — managing competing priorities, handling unexpected changes, being the person who knows how to get things done in a specific organisation — still requires human judgment and institutional knowledge.

What Stays Human

What to Do Next

  1. Develop into executive assistant or chief of staff roles. Senior EAs who support C-suite executives — managing complex stakeholder relationships, handling confidential strategic matters, and serving as a trusted extension of the executive — are providing high-value professional support that is significantly more resilient than general administrative work. This transition requires building deep trust and expanding into strategic coordination.
  2. Build expertise in specific business operations areas: project coordination, HR operations, or office management. Administrators who develop specialist knowledge in a function — who become the expert in a specific system or process — are significantly more valuable than those who remain generalists. PRINCE2 Foundation, CIPD Level 3, or specialist system certifications are concrete paths.
  3. Become a power user of AI productivity tools. The administrative professional who can use AI tools to handle more work with less time — and who understands the organisation's workflows well enough to identify where AI can add value — is positioning as an operations coordinator rather than a task executor. This reframing from "the person who does tasks" to "the person who makes things work" is the key career shift.
Sources: O*NET Online (onetonline.org) · ESCO (esco.ec.europa.eu) · All task data cross-referenced against O*NET occupation profiles. This analysis uses task-level exposure, not occupation-level prediction.