Samuel Thompson

AI investment is shifting focus from broad enthusiasm to essential data center infrastructure. Investors are prioritizing companies with substantial computing power and facilities, as highlighted by Goldman Sachs’ “flight to quality.” AI workloads will significantly boost data center capacity and energy demand, with infrastructure limitations now shaping AI strategy and geographical site selection. This pragmatic phase emphasizes foundational hardware and energy solutions over experimental software.

The U.S. Treasury Department, in collaboration with the Cyber Risk Institute and over 100 financial institutions, has launched the Financial Services AI Risk Management Framework (FS AI RMF). This framework provides a structured methodology for financial institutions to identify, evaluate, manage, and govern AI risks. It complements existing frameworks by offering sector-specific controls and practical guidance to address unique challenges like algorithmic bias and cybersecurity vulnerabilities. The FS AI RMF integrates AI governance into existing GRC processes, offering tools to assess AI adoption stages and implement robust control objectives across functions like Govern, Map, Measure, and Manage, promoting trustworthy and responsible AI deployment.

NTT DATA launches an “enterprise AI factory” initiative, leveraging NVIDIA’s GPU-accelerated platforms and software. This solution bridges the gap between AI pilot projects and production deployments, offering a repeatable blueprint for scaling agentic AI. It integrates NVIDIA NeMo and NIM Microservices for a full-stack platform deployable across cloud and edge. The offering aims to standardize AI outputs, reduce time-to-value, and drive measurable returns, as demonstrated by real-world deployments in healthcare, automotive, and manufacturing.

OpenAI’s Frontier platform challenges the traditional software revenue model by acting as an enterprise AI agent semantic layer. It integrates disparate systems, providing AI coworkers with comprehensive business context. This approach aims to reduce fragmentation and improve efficiency, with early adopters reporting significant time and cost savings. Frontier’s open architecture supports agents from various providers, disrupting per-seat licensing models and forcing incumbents like Salesforce and ServiceNow to adapt their pricing and strategies. The core debate is whether AI agents should be embedded or operate as an overlay intelligence layer.

E.SUN Bank and IBM have developed a comprehensive AI governance framework for the financial sector, addressing critical challenges of model validation, accountability, and regulatory compliance. This framework adapts global standards like the EU AI Act and ISO/IEC 42001, offering banks a structured approach for pre-deployment reviews, ongoing monitoring, data utilization, and risk assessment. The initiative aims to empower financial institutions to scale AI adoption confidently while ensuring robust oversight and regulatory adherence.

AI is transforming cybersecurity, accelerating threats and creating new challenges. AI security platforms in 2026 focus on securing enterprise AI usage, protecting AI models and infrastructure, and defending against AI-powered attacks. Leading solutions from Check Point, CrowdStrike, Cisco, Microsoft, and Okta offer distinct strengths in unified security, endpoint integration, network visibility, signal scale, and identity governance, respectively, to address this evolving landscape.

Physical AI, specifically humanoid robots, are proving vital for customer service, addressing labor shortages and the limitations of traditional automation. A partnership between KDDI and AVITA exemplifies this, combining robust communication infrastructure with advanced avatar technology. These robots, designed with human-like expressions and movements, aim to replicate nuanced nonverbal cues crucial for hospitality. Leveraging powerful data centers and AI models like Google’s Gemini, they are being prepared for commercial trials, marking a significant step from digital to physical customer interaction.

Banco Santander and Mastercard successfully conducted a live, AI-initiated payment within a regulated European network. This “agentic payment” trial, using Santander’s live infrastructure and Mastercard Agent Pay, demonstrated an AI agent autonomously completing a transaction within defined limits and controls. While not a public service, the pilot validates the technical feasibility of AI managing payments securely and compliantly, paving the way for future autonomous financial operations.

MWC 2026 showcased tangible AI-RAN progress, moving beyond discussion to field trials and product launches. Nvidia announced a major coalition committed to building AI-native 6G, releasing open-source AI tools. Nokia demonstrated successful AI-RAN tests with operators, integrating AI and RAN workloads on shared hardware. Ericsson presented an alternative strategy with its own AI-ready silicon. Operators like SK Telecom and SoftBank revealed plans for AI-native infrastructure overhauls. The hardware ecosystem is also maturing with new AI-RAN products. These developments signal a fundamental shift in connectivity delivery and monetization.

SK Telecom is undergoing a sweeping AI transformation, rebranding its entire operation as “AI Native.” This strategy involves overhauling IT systems, building gigawatt-scale data centers, and enhancing its large language model to over a trillion parameters. The company aims to offer hyper-personalized services, autonomous network operations, and a unified AI agent for seamless customer experiences, positioning South Korea as a global AI leader.