Slide showing where AI has been built into the ArcGIS platform
IGIS
Article

2026 ESRI User Conference – Day 2

If I had to summarize Day 2 of the ESRI User Conference in two letters, it would be AI. This year’s conference theme is “GIS – Creating a More Intelligent World,” and it is clear that ESRI means that quite literally. Artificial Intelligence was woven into nearly every session I attended today, and ESRI is making the case that AI is not a bolt-on feature but a fundamental part of how ArcGIS will work going forward.

ESRI is framing their AI work in three broad areas, each with its own focus. The first is AI Tools & Models, which they describe as advancing the science of GIS. These are the deep learning models, imagery analysis tools, and pretrained geospatial models that let us extract features, classify land cover, and analyze imagery at scales that simply were not practical before. The second is AI Assistants, aimed at productivity and enhancing the user experience. Think of these as helpers embedded throughout ArcGIS that let you ask questions in plain language, get help with workflows, and author maps and analyses more quickly. The third, and perhaps the most forward-looking, is Agentic AI, focused on impact by automating tasks and expanding what GIS can do. Rather than just assisting a user, agents can carry out multi-step geospatial workflows on our behalf.

What ties all of this together is that it is built on the ArcGIS system itself, the GIS tools and data, agents, models, and large language models all working within the platform we already use. That matters, because it means the AI is grounded in authoritative geospatial data and the analytical tools we trust, rather than a general-purpose chatbot guessing at spatial questions.

As GIS professionals, I think we will all need to spend some time considering what these capabilities mean for our own organizations and workflows, both the opportunities and the questions they raise. In the coming days I am looking forward to seeing more of these tools demonstrated in the technical sessions and getting hands-on where I can. I will write more about the advancements as the week progresses.