Administrative Assistant

Observed AI exposure is 45.28%, and modeled median job-loss risk is 10.03%. The weak point is the task mix: scheduling, correspondence, records, meeting notes, document formatting, travel support, and routine caller routing are all software-reachable. Local office context, confidentiality, and gatekeeping protect more of the job than routine customer-service queues, but scheduling, correspondence, records, and routing remain highly exposed.

AI can help an assistant draft messages, summarize meetings, schedule, prepare documents, organize files, and track follow-up. The upside mostly helps the employer handle office support with fewer hours or fewer people. A worker benefits more when tools free time for confidential coordination, office operations, vendor follow-up, or executive-support judgment.

The job is mostly office and screen work. Federal physical data shows light lifting, limited standing or walking, very low outdoor work, and almost no hazardous exposure. Those conditions add no physical barrier against substitution.

There is no occupational license, board exam, or protected scope for broad administrative-assistant work. The Certified Administrative Professional credential can signal seriousness, but it is voluntary and does not stop employers from automating routine office tasks.

Physical robotics is not the replacement channel for this job. The active pressure comes from software: calendars, email, document tools, meeting transcription, workflow systems, and AI assistants. Robot deployment does not change the office-support score.

The entry path is short: high school is usually enough, followed by short-term employer training. Office software skill, discretion, and local knowledge can build value, but the broad occupation does not require a long portable credential ladder.

The occupation is huge, with about 1.94 million jobs and about 202,800 annual openings, but employment is projected to decline about 1.6%. Because the job base is shrinking, the openings are treated mainly as replacement hiring rather than a strong growth signal.

The demand source is mostly replacement in a declining office-support occupation. Offices still need trusted coordination, but the broad need for routine scheduling, correspondence, filing, and document support is being reduced by software and shared tools.

Resilience is limited because employment decline and active office automation point in the same direction. The remaining human value is local coordination, confidentiality, and trust. That value can protect specific roles, but it does not rescue the broad occupation from shrinking routine support needs.