EXPOSURE TO AI
40%
MODERATE
OBSERVED IN REAL USE · Anthropic 2026
n/a
Not yet measured in the Anthropic Economic Index. The exposure figure is a capability estimate only.
A co-pilot is coming for the busywork, not the wheel.
40% of this role’s O*NET tasks are within reach of today’s AI. That is the core-weighted exposure score from Eloundou et al. 2023 (“GPTs are GPTs”). It measures a capability ceiling, not a headcount forecast. In the blast radius. A real slice of the work is already automatable. The rest isn’t.
WHAT AI CAN ALREADY DO
- Develop and document security procedures, policies, or standards
- Explain company policies and procedures to staff using oral or written communication
- Log items distributed to persons, such as keys and key cards
- Write reports documenting observations made while on patrol
- Order materials or supplies, such as keys, uniforms, and badges
WHAT IT STILL CAN’T
- Apprehend or evict trespassers, rule violators, or other security threats from the premises
- Call police or fire departments in cases of emergency, such as fire, bomb threats, and presence of unauthorized persons
- Inspect and adjust security equipment to ensure it is operational or to detect evidence of tampering
- Investigate disturbances on the premises, such as security alarms, altercations, and suspicious activity
- Monitor and authorize entry of employees, visitors, or other persons
THE HONEST PART. A percentage is not a pink slip. High exposure usually means a role shrinks and shifts toward judgment, direction and responsibility: the parts a model can’t sign its name to. Exposure ≠ displacement. Breathe.
"My job is 40% cooked. What’s yours?"
SOURCES: O*NET 30.3 occupational tasks · Eloundou et al. 2023 (“GPTs are GPTs”,
arXiv:2303.10130) · Anthropic Economic Index 2026 (CC-BY) |
how this is calculated |
last updated 2026-07-16