Cartographer and Photogrammetrist

Cartographer and Photogrammetrist scores 22/100 on this site's structural exposure measure, and unlike several other roles reviewed in this industry, the deeper research doesn't rescue that reading. The occupation's core task — converting raw aerial, satellite, or drone imagery into finished maps and structured GIS data — is exactly the kind of image-to-structured-data conversion that deep learning has gotten good at fast.

Esri, the dominant GIS software vendor, now markets AI-assisted feature extraction directly inside ArcGIS Pro, letting an analyst extract building footprints, roads, and land-cover boundaries from satellite or drone imagery in minutes rather than hand-digitizing them.

Microsoft has published its own deep-learning pipeline for extracting building footprints from satellite imagery at country scale, and a growing body of 2025-2026 academic research on convolutional and transformer-based models for building and parcel extraction confirms this is an active, maturing capability rather than a lab curiosity.

The exposure isn't evenly spread across the occupation's two halves, though this site's methodology scores it as one combined O*NET code. The tools above automate the photogrammetrist side most directly, extracting measurements and features from imagery, work that's inherently a structured, repeatable image-to-data conversion.

A cartographer's finished-map judgment, choosing what to symbolize, generalize, or emphasize for a specific audience and use case, sits a step further from what these tools do well today, though that distinction narrows every year as feature-extraction output requires less manual cleanup before it's map-ready.

Licensing offers essentially no protection nationally. O*NET and BLS list no state license requirement for the occupation as a whole, and while Virginia, Oregon, and North Carolina do license a narrow 'surveyor photogrammetrist' credential, it applies specifically to using photogrammetric methods to determine legal boundaries and topography — work that already falls under land surveying's own licensing regime — not to the general cartographic and GIS map-production work most of this occupation's roughly 14,700 workers actually do day to day.

There's also no equivalent to a surveyor's signed, sealed plat: a finished map or GIS layer doesn't require any individual's personal legal sign-off before use.

The work is also mostly desk-based rather than hands-on: BLS describes the job as spending 'much of their time in an office setting,' with fieldwork limited to occasional validation trips, so there's little of the physical-dexterity protection that shields a land surveyor's actual measurement work.

Two genuine, disclosed counter-signals keep this from reading as an unqualified decline.

BLS projects 7% employment growth through 2035 (a 'Bright Outlook' occupation) on rising demand for the underlying geospatial data, as drone mapping, disaster response, autonomous-vehicle mapping, and climate monitoring all consume more raw imagery every year, and one Amsterdam-based digital-mapping company's mid-2025 workforce cuts — though concentrated in its apps and sales divisions rather than core cartography — show at least one major mapping employer explicitly reallocating headcount toward AI rather than away from it entirely.

A 2025 academic review of generative AI in map-making also found current AI tools still fall short of professional cartographic quality working alone, real evidence against a fully hollowed-out occupation today, though not evidence of a legal or physical wall protecting it going forward.