On April 28, SAS announced platform updates focused on governance for agentic AI: policy controls, traceability, and accountability hooks built into the platform itself. The framing is the news. SAS is now leading with governance, not capability.
That tracks with what operators report. Roughly 10% of enterprises have moved agents past pilot. The other 90% are stuck on the same set of questions. Who is accountable when an agent resolves an exception incorrectly. What evidence exists to replay the decision. Where the human stays in the loop, and when. Until a vendor can answer those clearly, agents stay in pilot. Governance is not a phase that follows scaling. It is what makes scaling possible at all.