The UK's decision not to build one horizontal AI law has not removed regulatory uncertainty from corporate planning. British Business Review's October 2026 audience research suggests that uncertainty is already affecting project timing.

Across 4,189 responses, 34.1% said uncertainty around AI regulation had delayed an AI project. The sample is self-selected and not representative of all UK businesses, but it provides a useful measure of friction among a readership actively engaged with technology and business policy.

Copyright was the most commonly selected individual area of uncertainty. That aligns with the continuing policy debate. The UK government published its report and impact assessment on copyright and AI in March 2026 and stated it would not introduce reforms until it was confident they met economic and public-interest objectives.

For companies training, fine-tuning or procuring generative systems, uncertainty can affect data licensing, model selection, documentation and the willingness to deploy tools into workflows where copyrighted material is central.

Cyber risk is becoming part of the AI adoption budget

45.1% of respondents said they had experienced an AI-assisted cyber incident or attempted attack in the previous 12 months, while 86.2% expect cybersecurity spending to increase. Those audience findings sit beside a wider official threat environment in which 43% of UK businesses told the government's 2025/2026 Cyber Security Breaches Survey they had identified some form of breach or attack in the previous year.

The government survey also found only a minority of organisations using or considering AI already had cyber-security practices specifically designed for AI risk. That gap helps explain why security spending can rise alongside adoption.

The market for AI security is already expanding

The government's 2026 cyber-security sector analysis estimates 111 UK-registered firms explicitly offering AI-security products or services, up 68% from the previous baseline. AI/ML model security, advisory work and runtime or infrastructure security were the most common offerings.

For businesses, the practical implication is that AI deployment and cyber procurement are converging. Security is becoming part of the application architecture rather than a separate control added after launch.

Methodology

The October 2026 findings come from 4,189 British Business Review audience responses. The sample is not weighted to the national business population and should not be read as representative polling.

BBR will use repeat waves to track whether regulatory delay falls as policy becomes clearer and whether cyber spending expectations translate into actual budget growth.