© Redbow Consulting Group 2026, All rights reserved External Reference The Report noted that stringent but sometimes cumbersome governance frameworks can lead to shadow AI adoption and slowed deployments: External sources extend this discussion by advocating progressive, risk - calibrated governance models that scale oversight proportional to agent autonomy and contextual complexity. The notion of layered control through human - in - the - loop (HITL) for high - risk tasks and human - on - the - loop (HOTL) for routine autonomy, combined with continuous auditing and ‘guardian’ supervisory agents, constitutes a new governance frontier. This balance maintains regulatory compliance and cybersecurity w hilst enabling innovation velocity — a tension explicitly acknowledged in the Report’s risk versus innovation theme. The imperative for cross - functional governance bodies involving legal, medical, IT and compliance stakeholders aligns with the comprehensive frame works emerging from WEF - Capgemini reports. Recommendations: • Establish comprehensive, enterprise - wide data governance frameworks standardising data formats and taxonomies to promote consistency and seamless interoperability. • Introduce dedicated AI stewardship roles tasked with expediting approvals and preserving governance agility, ensuring alignment with the rapid evolution of AI technologies. • Reinforce data tagging requirements through incentive schemes that motivate and reward compliance, fostering organisational - wide accountability. • Build secure, user - friendly internal AI platforms to curtail shadow IT and personal device usage, thereby strengthening data security and governance control. • Form cross - functional AI governance bodies comprising legal, medical, IT, compliance, and operational leadership to ensure robust oversight and cohesive decision - making. • Engage legal and compliance professionals early in the AI lifecycle to clarify acceptable use cases and establish monitoring mechanisms that mitigate regulatory risks. • Prioritise pilot projects that deliver tangible business benefits with manageable regulatory exposure, progressively cultivating organisational trust. • Recognise AI adoption as fundamentally a human - centred process, ensuring operational managers are deeply engaged and, ideally, lead initiatives even when technical experts drive the technical implementation.