Loading role analysis…
Role summary
This role spans employment data, records, and first-draft communication as well as interviews and people decisions. The report classifies each work activity separately instead of treating the whole role as one automation decision.
Recruit, screen, interview, or place individuals within an organization. May perform other activities in multiple human resources areas.
Use this public role as a baseline. Compare it with the recurring work, tools, workload, controls, and exceptions in the real role before planning a workflow pilot or capability response.
OpenAI task exposure score
This score combines the OpenAI study's three task classes. E1 tasks count fully, E2 tasks count half because they require extra software or workflow changes, and E0 tasks count zero. It measures technical exposure—not the probability that the role will change.
33% E1 direct exposure + half of 62% E2 tool-dependent exposure = 64%
E1 · Direct exposure
33%
8 tasks
E2 · Needs AI tools
62%
17 tasks
E0 · Below threshold
4%
1 task
26 of 26 tasks have an exact OpenAI classification
26 tasks have at least one named research source attached. Missing classifications are not counted as low change.
This is not a job-loss forecast, a probability, a share of working hours, or a recommendation to automate the role. The E0/E1/E2 classes are shown beside every task below.
OpenAI beta exposure measure: E1 contributes 1, E2 contributes 0.5 because complementary software or workflow investment is required, and E0 contributes 0. Core tasks receive twice the weight of supplemental tasks, and weights are normalized across matched tasks. This is a technical-exposure index, not a probability.
Main source: OpenAI GPTs are GPTs (March 2023). Anthropic observed-use findings and other role-level studies answer different questions, so they remain separate and do not change this score.
No commercially usable, source-backed task-duration dataset currently covers these occupation tasks well enough. O*NET task frequency is shown as context but is not treated as work-time share.
Tasks to investigate
Use these classes to identify work worth closer investigation. E1 means the study threshold was met with a language model alone, E2 means it was met only with AI-powered software, and E0 means it was not met. They do not show adoption, work-time share, or whether an activity should be automated. Anthropic observed use stays separate.
26 of 26 work activities
OpenAI task exposure class
Direct exposure to language AI
OpenAI classified this task E1: a language model alone could reduce the time needed by at least 50% while keeping quality constant. This is a technical-exposure class, not a probability or an adoption forecast.
This class is not the probability that the task will change.
Similar work was also observed in professional AI use
OpenAI GPTs are GPTs
March 2023
OpenAI classified this task E1: a language model alone could reduce the time needed by at least 50% while keeping quality constant. This is a technical-exposure class, not a probability or an adoption forecast.
EconEvals June 2026
June 2026
Moderate modeled time saving. In the EconEvals study, a text-only chatbot was estimated to reduce the time needed for this task by 25-49%. This is a model-based research estimate, not observed workplace automation.
Limits to keep in mind
Anthropic Economic Index
Release date not recorded
Anthropic observed use matching this task in its public snapshot (published penetration value 0.910). This describes use in the source sample, not the share of workers using AI and not task automation potential.
These are survey results from the occupation data, not AI scores. Task frequency describes how often a task occurs, not how much work time it consumes.
Sources and context
The task list above shows the OpenAI E0/E1/E2 findings behind the exposure score. This section keeps observed use, occupation-wide studies, and other source records separate because they answer different questions.
These records describe the occupation as a whole. They are shown as separate context and do not change the task-based exposure score.
AI USAGE
ANTHROPIC
Mapped directly to this occupation.
medium
AI EXPOSURE
MICROSOFT
Mapped directly to this occupation.
low
AI EXPOSURE
OPENAI
Mapped directly to this occupation.
high
ROUTINE INTENSITY
OECD · GROUP MAPPING
low
These are the occupation records used for the role structure. Public AI evidence links are shown beside the matched task evidence above.
O*NET
View source recordOnly exact OpenAI E0/E1/E2 task classifications enter the role exposure score. Anthropic observed use, EconEvals modeled time saving, and role-level benchmarks stay separate because they answer different questions.
Task sources
Tasks
O*NET · O*NET Task Statements.txt · db_30_0_text
Missing a role?
Tell us which occupation you would like us to review for a future public analysis.
Request this roleYour real work
Use a pilot code, start from this occupation, and map the tasks you actually do. The complete report is included in exchange for short feedback.
Use a pilot code