Siemens and NVIDIA are building self-verifying agentic AI into electronic design automation (EDA) tools, part of a wider corporate push to deploy autonomous AI agents across core workflows.1 The rollout is moving faster than the internal rules meant to control it.
Microsoft is expanding infrastructure and Foundry/Copilot Studio integrations with Mistral, while insurer Manulife has widened its own Microsoft partnership specifically around AI governance. "Our partnership with Microsoft is a critical enabler of Manulife's continued evolution into a truly AI-driven organization," said Shamus Weiland, describing the Microsoft Frontier Suite as a foundation meant to "embed AI responsibly across the franchise."2
Snowflake has launched CoCo, an agentic control plane giving enterprises a governed environment to manage workflows across data, models and applications.3 Customers including Fanatics, Thomson Reuters and WHOOP are already using it to accelerate AI deployment at scale.3 Box has added new agent security controls to the same end: keeping autonomous systems inside defined boundaries as they touch sensitive company data.
The common thread is governance, not capability. Survey data shows the top obstacle to agentic AI adoption is security and privacy, not technical performance, and many leaders admit they still lack a clear roadmap or ownership model for the agents they are deploying.4 Prasun Shah, writing on organizational design in the agentic era, argues AI agents now function as "connective tissue" moving across technology layers rather than sitting in a single stack. "That is where the next battleground will be," he said.5
For European companies, that battleground overlaps directly with regulation. The EU's AI Act imposes risk-based obligations on automated systems making or influencing decisions, and agentic tools that act autonomously across data and applications sit closer to its higher-risk categories than static chatbots. Siemens and NVIDIA's emphasis on self-verification in EDA, and Snowflake's framing of CoCo as a governed control plane, point to vendors building compliance-adjacent features ahead of enforcement rather than after it. Enterprises adopting these tools now are effectively pre-positioning for audits, ownership disputes and liability questions that EU rules are set to formalize.
