AI is hot as volcano, and AI Agent is popping out as politicans' promises.
Any workable AI [Agent] with SAS, from basic data cleansing, data analysis/graphing, to optimization alike solution?!
In December 2025, SAS released its SAS Viya Copilot capabilities with the 2025.12 release. This includes an AI code assistant (in VS Code), model building and explanation in Model Studio, and also "augmented analytics" in SAS Visual Analytics (now in preview, not yet production).
SAS Viya Copilot takes advantage of the SAS Viya open architecture and APIs to provide AI-assisted insight and take actions on your behalf. These same APIs make it possible for other AI tools to also access and drive SAS capabilities.
For SAS 9 users who aren't yet using SAS Viya, I find that AI tools can be very helpful too. Many of them (like GitHub Copilot and Claude) are getting better at understanding and generating SAS programs. Although SAS 9 doesn't have as many open APIs as SAS Viya, those that it does support have been around for a long time, which means that AI tools have a lot of established documentation and examples to train on, and all can help SAS users get more done.
AI capabilities from SAS are delivered in SAS Viya and other modern fit-for-purpose tools like SAS Data Maker (for synthetic data). SAS is also a major player in the AI governance space, providing customers with a way to manage all of their AI models and capabilities in a way that mitigates risk and bias, and prepares for emerging compliance initiatives in different regions.
If you really want to learn more, be sure to register for SAS Innovate 2026! Full agenda will be published soon, and you'll see a lot of Agentic AI and related topics featured.
I just spent and hour or two vibe-coding some simple analysis and graphing using VS Code with the SAS extension and GitHub Copilot , connected to my local PC SAS (yes still have it!) and was blown away by what it can do.
I asked it to make a plot using sashelp.baseball (inspired by Joe's SAS Innovate presentation), of at bats vs batting average, with a regression line overlaid, with an image of a baseball in the background, followed by a PROC REPORT step listing the top 10 players (by batting average), highlighting the players who hit over 350.
Crazy stuff. First time ever vibe coding, and I can see what all the fuss is about.
The agent would write the code, batch submit it, fix the code, batch submit again....
It even scanned the log to look for errors, without me asking!
Nearly 200 sessions are now available on demand in the Innovate Hub.
Watch Now →Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.