Three-Quarters of UK IT Leaders without Strong AI Governance Plans

Why It Matters

Without clear governance, firms risk vendor lock‑in, compliance breaches, and eroding trust, while regulators scramble to catch up. The gap could shape future policy and competitive dynamics in Europe’s AI market.

Key Takeaways

• 87% use agentic AI; only 25% have strong governance.
• Only 67% have an AI exit strategy, risking vendor lock‑in.
• 89% call for government AI sovereignty policies and open‑source rules.
• By 2026, just 48% will know where AI data resides.
• Governance gap hampers visibility between AI deployment and control.

Pulse Analysis

The Red Hat survey underscores how quickly agentic AI has moved from aspiration to reality for UK enterprises. While 87% of IT leaders report active use, only 25% have instituted comprehensive governance, leaving a sizable control gap. This mismatch fuels uncertainty around data stewardship, model provenance, and compliance, especially as AI models become more autonomous and integrated across critical business processes. The lack of oversight not only threatens operational stability but also exposes firms to regulatory scrutiny as governments tighten AI oversight.

European regulators are feeling the pressure, and UK decision‑makers are vocal about the need for AI sovereignty. Nearly nine in ten respondents want legislation that enforces open‑source principles and mandates transparent, auditable AI practices. The drive for sovereignty reflects concerns over the dominance of US‑based AI providers, which can limit data locality and increase dependency on foreign infrastructure. Open‑source mandates are seen as a pathway to regain control, foster innovation, and ensure that critical AI workloads can be audited or relocated without prohibitive costs.

For businesses, the practical implications are immediate. With only 67% having an exit strategy, many risk vendor lock‑in that could disrupt operations if a provider changes terms or technology. The projected visibility gap—just 48% will know where their data resides by 2026—highlights the urgency for robust data‑mapping and governance tools. Companies that proactively build modular AI architectures, adopt open‑source models, and negotiate clear exit clauses will be better positioned to navigate the evolving regulatory landscape and maintain trust with customers and stakeholders.