- Snowflake Cortex Code has become the fastest-growing product in Snowflake’s history, according to the company, with customers including Block and Deloitte deploying Claude models directly on Snowflake-hosted data, a sign that governed, production-ready AI is moving from pilot to standard infrastructure.
- Microsoft Power BI’s June 2026 Copilot update lets users modify semantic models and build operational apps via natural language, compressing data cleanup tasks that previously took hours into a process the company says takes minutes, though independent validation of those time savings is limited.
- Palantir’s expansion with GNP Seguros deploys AI agents to detect claims fraud and sharpen underwriting across all lines of business at Mexico’s largest insurer, while a separate June 22 announcement confirmed Palantir’s role in the U.S. Army’s NGC2 command and control data layer.
- Databricks’ Genie One, now generally available after the Data + AI Summit 2026, automates reporting and task execution for business teams in marketing and finance, but its real-world utility depends on the quality of the governed data foundation underneath it.
- Google Cloud Looker’s June 2026 updates embed Gemini directly into Explore, with AI-generated queries and an Insight Assistant that modifies visualisations via natural language, shifting Looker from a reporting tool toward one that can surface analysis without a dedicated analyst.
Snowflake, Microsoft, Palantir and Google all made substantive AI announcements in June 2026, and the through-line is the same: governed data as the prerequisite for production AI. The era of sandboxed experiments is giving way to multi-year enterprise contracts, army-level deployments and analytics platforms that business users can drive without writing a query.
Snowflake Cortex AI Accelerates Production AI
At Snowflake Summit 2026 in early June, Snowflake and Anthropic reported strong momentum in their strategic partnership, centred on deploying Anthropic’s Claude models through Snowflake Cortex AI. The pitch to enterprise customers is straightforward: run AI directly on Snowflake-hosted data without moving it, preserving existing security and compliance controls. Customers including Block and Deloitte are among those using the combined platform for financial analysis and cybersecurity investigations, according to the companies. Snowflake claims Cortex Code has become the fastest-growing product in its history, driven by demand for AI-powered coding agents inside the Snowflake environment. The broader Cortex AI integration with Claude builds on an expanded partnership announced in December 2025, with the aim of helping enterprises move operational workloads, not just pilots, onto the platform.
Microsoft Power BI Copilot Transforms Data Modeling
The June 2026 update to Microsoft Power BI added a set of AI capabilities that, taken together, push the product toward a model where business users can interact with data infrastructure that previously required dedicated analysts or data engineers. The headline feature is Copilot in Web Modelling, currently in preview, which lets users modify semantic models through natural language: reviewing for inconsistencies, renaming tables and columns, building relationships and generating DAX measures. Microsoft says this compresses hours of manual cleanup into a rapid review process, though independent validation of those time savings is limited. Fabric Apps for Semantic Models, also in preview, offers an AI-first path for building operational applications on top of existing semantic models, designed so that business logic and governance are inherited rather than rebuilt. Separately, Fabric IQ’s integration into Microsoft 365 Copilot Chat lets business users query Power BI data directly from their existing productivity tools, bypassing the need to open Power BI for routine questions.
Palantir Foundry and AIP Drive Operational Value
In June 2026, Palantir announced an expansion of its contract with GNP Seguros, Mexico’s largest insurer, to deploy its Artificial Intelligence Platform (AIP) across all lines of business including health, life, auto and damage insurance. The initial deployment focused on claims fraud detection; the expanded scope adds AI agents that continuously monitor risk, sharpen underwriting precision and process claims across GNP’s full operational data layer. The practical architecture here is Palantir’s Ontology: GNP’s claims, underwriting and operations data is unified into a single coherent structure, making it possible to identify anomalous claim patterns before payments are disbursed. That is a concrete fraud-prevention use case, not a proof of concept. On June 22, 2026, Palantir also confirmed its role in establishing the U.S. Army’s Next Generation Command and Control (NGC2) common data layer baseline, a separate engagement that speaks to the platform’s traction in defence and intelligence.
Google Cloud Looker with Gemini Empowers Business Users
On June 16, 2026, Google Cloud announced a set of AI updates to Looker Explore that embed Gemini more deeply into the analytics workflow. The Insight Assistant lets users modify data tables and visualisations through natural language, using the Conversational Analytics API to identify relevant fields and apply filters. The AI-assisted Quick Start addresses a persistent friction point: users arriving at an empty Explore interface with no obvious starting query. The feature generates queries automatically, drawing on Gemini to surface relevant fields beyond what is immediately visible. Google also added AI-generated Explore summaries and the ability to publish Looker BI Agents directly into Gemini Enterprise. The aggregate effect positions Looker as a platform where business users can reach analytical conclusions without routing requests through a data team, a meaningful change for organisations where analyst capacity is a bottleneck. For more on how AI is reshaping enterprise analytics infrastructure, see our coverage of Cognizant and KPMG’s multi-agent AI deployments.
Databricks Lakehouse Platform Unifies Data and AI
The Databricks Data + AI Summit 2026, held June 15-18, centred on a single argument: agentic AI only works if the data underneath it is governed and real-time. The most concrete announcement was the general availability of Genie One, an AI agent designed for business teams in functions like marketing and finance. Genie One automates task execution and report generation with what Databricks describes as full business context, though how well that holds up across heterogeneous enterprise data environments remains to be seen in practice. On the security side, Databricks introduced Automatic Identity Management (AIM) for Entra ID and new Context-Based Ingress policies for governing access to AI experiences, capabilities aimed at teams that need to scale AI usage without losing control over who can query what. Databricks also has DBRX, its open-source large language model, integrated across its GenAI products; the company has pointed to strong benchmark performance in SQL and retrieval-augmented generation tasks, though the more relevant measure for enterprise teams will be how it performs on their own private data. The broader pattern across all five platforms covered here is consistent: the infrastructure investment is shifting from model selection toward data governance. For enterprise teams evaluating where to focus, that framing matters more than any individual feature announcement. For more analysis on enterprise AI strategy, visit our Enterprise AI section.
