Databricks rolled out a series of security and compliance features at Data + AI Summit 2026 on June 19, targeting the friction that keeps regulated industries from scaling AI. The updates include General Availability of Automatic Identity Management (AIM) for Microsoft Entra ID on AWS and GCP, Public Preview for Okta, context-based ingress policies, and a new Private Network Gateway. The company also expanded its compliance portfolio with HITRUST, ISMAP, and upcoming FedRAMP High support. These moves arrive as 62.1% of organizations now consider AI-powered defensive tools a necessity, according to Futurum Group's 2H 2025 Cybersecurity Decision Maker Survey of 1,008 decision-makers.
The platform updates span all three major clouds-AWS, Azure, and Google Cloud-and target the operational realities of multi-cloud AI workloads. Context-based ingress policies enable zero-trust access to Genie, dashboards, and AI applications. Expanded Private Link support now covers Lakebase and account-level resources, tightening network isolation for serverless environments. The Private Network Gateway aims to simplify secure connectivity without forcing teams into complex network re-architecture.
Identity management shifts from bottleneck to enabler
Automatic Identity Management for Entra ID and Okta removes manual provisioning steps that slow down AI access. As organizations move beyond pilot projects and open data tools to larger user bases, the risk of permission drift multiplies. "62.1% of organizations now view AI-powered defensive tools as a necessity," said Futurum Group's survey data, reflecting a shift toward automated, scalable identity controls. The real test will be whether Databricks can integrate AIM across multiple identity providers and clouds without introducing new attack surfaces or configuration gaps.
Zero-trust access and private connectivity become mandatory
Context-based ingress and expanded Private Link support signal a departure from static perimeter defenses. The move toward dynamic, context-aware controls aligns with what enterprises need as AI workloads shift to serverless and operational environments. The Private Network Gateway offers high-throughput connectivity without requiring teams to rebuild their network architecture. This matches a trend Futurum identified earlier in 2026: organizations are abandoning custom architectures in favor of validated reference designs. The challenge remains maintaining granular policy enforcement as data access decentralizes across hybrid estates.
Compliance expansion targets regulated industries directly
Databricks' updated compliance certifications-HITRUST, ISMAP, and the planned FedRAMP High on Azure Commercial-address the barriers that have kept healthcare, government, and financial services from adopting cloud-native AI at scale. Offering consistent controls across AWS, Azure, and Google Cloud is a differentiator for enterprises that want to modernize without increasing risk exposure. However, as regional compliance frameworks diverge and new mandates emerge, the operational burden of maintaining certification parity across clouds will intensify. Databricks must prove it can keep pace with evolving standards while supporting rapid AI development cycles.
Why this matters for management
The security and compliance upgrades directly affect how fast teams can deploy AI without triggering audit findings. For managers in regulated industries, the shift from manual identity provisioning to automated controls reduces a recurring source of project delays. The expanded private connectivity options mean serverless AI workloads can connect to existing data sources without a six-month network redesign. The compliance certifications provide a clearer path for moving workloads that handle sensitive data into cloud environments. The decision point for management is whether these features work reliably across their specific multi-cloud setup-and whether the platform can sustain that reliability as frameworks evolve.
Your membership also unlocks:
