Adapting to AI at work

When a team starts using AI tools, the way it works starts to change. A few working habits make that change easier to follow and to check.

Understand the current process first

The government's AI Playbook, written for civil servants, suggests looking at the work as it is done today. It says "a user researcher can work with a business analyst to find efficiencies in business processes and measure dissatisfaction. This will help establish a baseline of metrics to measure your project outcomes against."

A smaller organisation may not have those roles. The same step still helps. Someone watches the work, writes down how it goes now, and notes what people find slow or frustrating.

Start with one small pilot

The government's Service Manual is written for teams building public services. It describes an early trial phase called alpha. It advises doing "the minimum you need to test your riskiest assumptions".

The same idea suits a first AI pilot. Pick one piece of work. Decide what good enough looks like before you start, then try the tool on it for a set time. The AI Playbook says "You must be clear about what objective measures you're assessing the AI outputs against".

Decide how you will measure it

HM Treasury's Magenta Book is government guidance "on what to consider when designing an evaluation". It includes a separate guide to evaluating the impact of AI.

Keep a register of the AI tools in use

Public bodies have a model to follow. The Algorithmic Transparency Recording Standard "establishes a standardised way for public sector organisations to publish information about how and why they are using algorithmic tools". Any organisation can keep a simpler list of its own, with columns like these:

The tool
What it is used for
Who owns it
What data it sees
When it was last reviewed

The government also consulted on a self-assessment tool for AI management, called AI Management Essentials. In its response, published on 6 February 2026, it decided to drop the tool. It plans simpler guidance for small and medium-sized businesses in its place.

A short policy staff can follow

The National Cyber Security Centre (NCSC) says "Managers don't need to be technical experts, but they should know enough about the potential risks from AI to be able to discuss issues with key staff." In an earlier blog post about chatbots such as ChatGPT, the NCSC advised "not to include sensitive information in queries to public LLMs". LLMs are the large language models those chatbots run on. That one line can open a staff policy. The Information Commission (which took over from the ICO on 30 September 2026) has guidance on AI and personal data.

A person checks what goes out

Principle 4 of the AI Playbook reads "You have meaningful human control at the right stages". In everyday work, a person reads the draft, the reply or the document before it leaves the organisation.