# AI Training for Business Teams

Source: https://dijituldevelopments.co.uk/ai-training-for-teams/
Updated: 2026-10-18

> dijitul developments runs practical AI training for UK business teams: safe, effective use of tools like ChatGPT, Microsoft 365 Copilot and Claude on your real tasks, clear data handling rules, and hands-on API workshops for developers. Sessions are built around your work, not generic demos, and quoted at a fixed price after a free chat.

## Problems this solves

- Some staff use AI daily and others are afraid to touch it
- People paste customer details into whatever free chatbot they find
- You bought Copilot licences and usage has flatlined
- Output gets sent to customers without anyone checking it
- Your developers want to add AI features but have never used a model API

## Key facts

- Built around your own documents and tasks, with sensitive data replaced by realistic samples
- Separate tracks for general staff, managers and developers
- Covers what AI gets wrong: invented facts, outdated knowledge and prompt injection
- Includes a plain-English AI acceptable use policy for your business
- Helps evidence AI literacy, which the EU AI Act has required of providers and deployers since 2 February 2025
- Taught by developers who build AI systems, on site or over Teams

## Training built on your work

Generic AI courses show someone writing a poem. Your team needs to draft a quote, summarise a tender, answer a supplier email or check a contract clause. Before training, we survey which tools you have, what jobs people do and how confident they are. We then build exercises from realistic versions of your own documents, with personal and commercial details swapped for safe samples, so people leave with methods they will use the next morning.

Sessions run on site or over Microsoft Teams, in small groups so everyone gets hands-on time. We are developers who build AI systems for a living, so the questions about how it works and why it went wrong get real answers.

## Tracks for different roles

- **Everyone.** What large language models are good and bad at; writing clear instructions with context, examples and a defined output; checking answers; using AI with documents and spreadsheets; what never goes into which tool.
- **Managers.** Spotting good uses in their team, reviewing AI-assisted work, measuring time saved, and handling the people side of change.
- **Power users.** Building reusable assistants: custom GPTs in ChatGPT, Projects in Claude, and agents in Microsoft 365 Copilot, grounded on approved documents.
- **Developers.** Calling model APIs, structured output, tool calling, retrieval, evaluation and guarding against prompt injection, as used in our LLM application builds.

## Safe use, not just clever prompts

The biggest risks with AI at work are ordinary ones: personal data typed into a consumer tool that may keep it, confident answers that are wrong, and AI-written text sent out unchecked. Training covers which tools are approved and why, the difference between consumer and business plans on data retention and model training, and simple checking habits for anything that leaves the building.

We draft a plain-English acceptable use policy with you, covering approved tools, data categories, human review and how to report problems. If you trade in the EU, the EU AI Act requires providers and deployers to take measures to ensure a sufficient level of AI literacy among staff who use AI systems, a duty that has applied since 2 February 2025. Training records and the policy help you show what you have done.

We teach checking as a habit rather than an afterthought. Staff learn to ask the tool for its sources, to compare a summary against the original on anything that matters, to spot answers that sound right but cite nothing, and to be suspicious of instructions hidden in documents or emails that an assistant has been asked to read. Developers go further, testing how their own features behave when someone deliberately tries to trick them.

## Getting the tools set up properly

Training works best when the tools are ready: business licences, single sign-on, data retention settings and permissions checked. Setting up and supporting Microsoft 365 Copilot and ChatGPT business accounts is IT administration, which our sister site dijitul support handles. If you have not yet checked what Copilot can see in your files, our AI readiness audit covers that.

We also check that the features we teach are switched on for your licences. It is frustrating to learn a technique and find the option missing, so we confirm the plan, admin settings and any connectors to SharePoint or other systems before the sessions, and adjust the content to what your team can actually use.

Single sign-on, retention settings and who can share custom assistants across the company all affect what we recommend, so we agree those points with whoever runs your IT before the first session.

## After the training

Every team gets a library of the prompts and workflows used in their sessions, written up for reuse. Around a month later we run a follow-up to see what stuck, answer questions from real use and add the best new examples to the library. When teams find a task that AI does well but that needs connecting to your systems, such as reading supplier invoices into Xero, that becomes a candidate for an AI automation build, and we can quote it. Training is quoted at a fixed price after a free chat, based on the number of sessions and tracks. See also AI consultancy, or get in touch.

## What we deliver

- A short pre-training survey of tools, tasks and confidence levels
- Workshops for staff, managers or developers, built on your tasks
- A prompt and workflow library for your common jobs
- A draft AI acceptable use policy
- Attendance and content records for your training files
- A follow-up session to answer questions after real use

Technologies: Microsoft 365 Copilot, ChatGPT Enterprise, Claude, OpenAI API, Anthropic API, Copilot Studio

## FAQs

### What does your AI training cover?

dijitul covers what AI tools are good and bad at, writing clear instructions, checking output, safe data handling and building reusable assistants. Developers get a separate track on model APIs, retrieval, evaluation and prompt injection. Exercises are built on realistic versions of your own tasks.

### Which AI tools do you train on?

Whichever your business uses or plans to use, typically Microsoft 365 Copilot, ChatGPT Business or Enterprise, and Claude. dijitul also explains the differences between consumer and business plans, especially on data retention and model training.

### Do you help write an AI policy?

Yes. dijitul drafts a plain-English AI acceptable use policy with you, covering approved tools, what data can go where, human review of output and how to report problems. It is part of the training scope.

### Does the EU AI Act apply to our training?

If you provide or use AI systems in the EU market, the EU AI Act has required measures to ensure staff AI literacy since 2 February 2025. dijitul training and its records help show the steps you have taken. Take legal advice on your specific obligations.

### Can you set up Copilot or ChatGPT for us?

Licensing, single sign-on and tenant settings are IT administration, which our sister site dijitul support handles. dijitul developments focuses on the training and on building custom AI integrations when a task needs one.

### How is training priced?

dijitul quotes a fixed price after a free chat, based on the number of sessions, tracks and how much tailoring your tasks need. The prompt library, policy draft and follow-up session are included in the quoted scope.

## Pricing and contact

Every project gets a fixed-price quote after a free initial chat and a short scoping stage. You own the code and the data. Book a free chat: https://dijituldevelopments.co.uk/contact/ · 01623 650333 · info@dijitul.uk