AI Resilience Report forAnthropologists and Archeologists

Right now, AI is mostly augmenting anthropologists and archaeologists rather than replacing them. In fieldwork, AI analysis of aerial surveys is helping archaeologists identify previously undiscovered Maya ruins beneath the jungle canopy, and AI-assisted surveys in Peru have doubled the known number of features in the Nazca Lines ^[1]. Researchers are also training deep-learning models on simulated archaeological sites to spot rare features — one recent test on LiDAR data from Louisiana's Kisatchie National Forest found all 12 known targets plus 11 additional objects worth further study ^[2].

LiDAR specialists note that archaeology is moving closer to true automation of feature detection in point clouds ^[3]. On the writing side, a large PNAS study covered by Inside Higher Ed showed scientists who lean heavily on AI to draft grant applications are more likely to get NIH funding, though possibly at the cost of novel ideas ^[4]. Hands-on tasks like excavation, mentoring students, and community consultation remain firmly human.

Adoption is uneven. Cheap, ready-to-use tools (ChatGPT, Claude, Mask R-CNN models, photogrammetry software) make image analysis and grant drafting easy wins, and shrinking research budgets push scholars to use them. But the field is cautious: the American Journal of Archaeology now prohibits using generative AI to write any part of a manuscript's text or notes ^[5], and anthropologists themselves are calling for a critical "AI anthropology" that studies how these systems reshape human culture rather than blindly adopting them ^[6].

Data quality is another brake — most archaeological datasets are incomplete, inconsistent, or in incompatible formats, while AI thrives on scale ^[1]. The takeaway for students: AI will handle more of the pattern-spotting and paperwork, but the human judgment, ethics, and fieldcraft at the heart of this career are exactly what employers still need.