Harvey Nash opens AI development to business teams
Harvey Nash is adopting Citizen AI, an approach that lets employees outside IT build tools, agents and prototypes within defined guardrails. The aim is to help staff test solutions to business problems before committing significant development resources, rather than create unofficial production systems. Technology teams then help proven concepts meet enterprise requirements for security, integration, testing, resilience and support.
A content manager has built an agent that helps recruitment consultants find approved marketing posts and assets, personalise them to their role and tone, and publish them on social channels. The tool supports brand amplification without requiring recruiters to create new content. Usage can be measured through the number of posts created, with gamification intended to encourage adoption.
Microsoft Copilot serves as the approved general-purpose enterprise AI environment. Other business applications require Group Technology approval. Controls cover sensitive information, data permissions, access to agent-building capabilities and consumption costs. More complex agents involving integrations require technology team support, while agents moving beyond personal use need a named owner responsible for their behaviour, data access and continued suitability.
CIO Ankur Anand emphasises that employees remain accountable for AI-assisted work. Staff need to understand data sensitivity, challenge outputs and recognise when human judgement is essential. Managers are expected to clarify acceptable data use and provide time to learn.