Figure 1
A flowchart titled 'Integrated Sociotechnical Exposure Framework (ISEF) Methodology Overview' depicts the process of ISEF. The flowchart starts with '1. Task Universe' which includes 674 tasks across 28 occupations. This leads to '2a. TA Variables' with seven forward indicators of technical suitability, and '2b. IR Variables' with five indicators of liability, privacy, and governance friction. These variables feed into '3. Scoring Protocol' using ChatGPT-5, DeepSeek-R1, and Qwen-3 with three runs each. The next step is '4. Aggregation' where importance-weighted task scores are aggregated to the occupation level. This leads to '5. ISEF Exposure' calculated as Exposure = TA x (1 - IR/100) reported as a descriptive indicator. The final step is '6. Robustness and Calibration' involving alternative functional forms plus preliminary practitioner calibration.Complete ISEF workflow for the 674-task, 28-occupation sample: task universe; TA and IR indicators; three-model rubric scoring; occupation-level aggregation; ISEF exposure calculation; and functional-form sensitivity plus preliminary practitioner calibration. Domain Expertise is treated as a TA variable, while accountability and sign-off barriers are captured through IR variables