What happened
KPMG and Anthropic announced on May 19-20, 2026, a global alliance embedding Anthropic's Claude AI directly into KPMG's Digital Gateway client delivery platform, with initial focus on tax and private equity clients. The integration allows clients to build agentic workflows in real time. KPMG is rolling out Claude access to all 276,000 employees globally and embedding Claude into PE-focused product offerings. The two organizations will co-develop new Claude-powered products for portfolio companies. KPMG stated the alliance embeds cybersecurity, risk, and AI assurance into how AI systems and tools are designed and operated, with teams finding and fixing vulnerabilities and protecting critical systems, guided by the KPMG Trusted AI framework. According to KPMG US Vice Chair for Tax Rema Serafi, building an AI agent to help clients adjust to changing tax regulations previously took weeks; with Claude Cowork and Managed Agents integrated in Digital Gateway, that capability now takes minutes.
Why it matters
The alliance represents one of the largest enterprise AI deployments announced in 2026, with a Big Four professional services firm embedding frontier AI into its core client-facing platform and making it available to its entire global workforce. The emphasis on tax and private equity as initial use cases highlights the shift toward domain-specific agentic workflows that require judgment, accuracy, and regulatory knowledge. KPMG's public commitment to embedding cybersecurity, risk, and AI assurance into the design and operation of Claude-powered systems signals that enterprise AI adoption is increasingly paired with governance frameworks and active vulnerability management. For AI security practitioners, the alliance demonstrates that large regulated firms are moving from pilot projects to production-scale AI integration, creating new audit, assurance, and compliance demand.
Applicability
Professional services firms, tax and legal advisory practices, private equity firms, and enterprises in regulated industries evaluating frontier AI for client-facing workflows. Security, risk, and compliance teams should note KPMG's stated approach of embedding assurance into AI system design rather than treating it as a post-deployment audit. Organizations considering large-scale AI deployments should evaluate whether their governance frameworks can support agentic workflows at the scale KPMG is implementing (276K users, client-facing).