Electrical work carries low displacement risk from generative AI — exposure concentrates in paperwork and diagnostics, not the tools

Asked (summary):

What does recent research say about how generative AI affects electricians and electrical trades workers specifically? Is electrical work at high or low risk of AI disruption, and what is the mechanism?

Five independent research or synthesis sources — three primary studies and two secondary occupational pages, each from its own host and URL, published between 2023 preprints and September 2026 — converge on one mechanism: generative AI works on information artifacts, not on walls, conduit or switchboards. A secondary page reporting Microsoft Research's measure puts U.S. electricians at about 15% AI applicability (roughly the 29th percentile of 342 occupations); a 2026 synthesis scores building electricians 31/100 exposure with a 0.42 occupation-specific GenAI risk score. These unlike metrics are shown side by side, never averaged.

Where generative AI touches an electrician's job — a task-layer spectrum

Information work — moderate exposure, mostly augmentationPhysical execution — low exposure, the substitution bottleneck

Positions are a qualitative placement on the exposure spectrum, derived from the task mechanisms described across the five sources; they are not measured scores. Hover a bar for the source reasoning.

What the visual cannot say alone

"Exposure" in GPTs are GPTs means an LLM could cut task time by at least 50% — it does not distinguish augmentation from displacement — ar5iv.labs.arxiv.org

The ILO's 2025 refined index says exposure usually implies job transformation rather than whole-job automation, and explicitly lowers scores for hands-on tasks such as welding and operating machinery — ilo.org

Microsoft's real-world Copilot study finds the highest AI applicability in creating, processing and communicating information, and the lowest in manual work, operating machinery and physically working with people — arxiv.org

RoleFate's 2026 synthesis for building electricians (31/100 exposure, 0.42 GenAI risk) attributes the score to AI help with drawings, circuit-test interpretation and fault triage — while installation, repair and accountable safety judgment keep the occupation in the hands-on-trades band — rolefate.com

Observed Claude usage underrepresents physical trades, but this measures Claude usage in a self-selected, technically skewed sample — not all AI use — pasqualepillitteri.it

Evidence matrix — five sources, one row each

Source	Type	Scope	Quantitative finding	Key caveat or mechanism	Date	Link

Method: audited evidence table of 5 rows, one per source, each backed by a distinct URL and host (no URL supports more than one row); dates span a 2023 LLM-exposure preprint through September 2026 syntheses. Quantitative findings are reported in each source's own metric (percent applicability, 0–100 exposure, 0–1 risk, ILO gradient) and are never averaged across sources. The task-layer spectrum is a qualitative placement built from the sources' task descriptions; long mechanism passages were shortened for space.

This report was generated automatically by Keenable SELECT at a user's request, from publicly available web sources linked herein. Keenable does not review, verify, or endorse its contents and makes no representation as to accuracy, completeness, or timeliness; AI-based extraction may contain errors. Nothing in this report is investment, legal, financial, or other professional advice. All trademarks and referenced content remain the property of their respective owners; no affiliation or endorsement is implied. To report an error, rights concern, or request removal: legal@keenable.ai.