ChatGPT workshop for companies: a working day for management and IT leadership
Everyone in your organisation uses ChatGPT, or will within six months. The question is no longer whether generative AI will break through, but whether you steer its adoption deliberately or let it happen by default. Our ChatGPT workshop for businesses is not a demo of prompt tricks, but a structured working day in which senior management, HR, L&D and IT management together draft an AI rollout plan, including a tool shortlist, governance, a training roadmap and monitoring.
Plan the working day View the programmeWhen 'everyone uses ChatGPT' can go well or wrong, and how a working day makes the difference
In most organisations, ChatGPT is already in use. Not through a formal decision, but through employees who use the free version privately for client emails, proposal text or code snippets. That brings productivity gains, and at the same time invisible risks: client data ending up in a training dataset, legal answers containing hallucinations, and an organisation with no idea which departments do or don't use AI. Shadow AI is the pattern, and it is rarely caused by bad intentions; usually an employee is simply filling a gap that management did not see.
The working day we facilitate brings senior management, HR/L&D and IT management to the same table to turn that situation around. Not by banning AI, which would lose you the battle because employees will use it anyway, but by designing a deliberate rollout in which the right tools, with the right licences, end up in the right place. By the end of the day you have no abstract vision, but a concrete rollout plan: which AI tooling you make centrally available, how you train employees, which data may and may not go into which model, and who is responsible for monitoring and adjustment.
At the same time, you meet an obligation that often goes overlooked: Article 4 of the AI Act requires organisations to ensure sufficient AI literacy among staff who work with AI systems. A documented workshop with follow-up is concrete evidence that you take that duty seriously, and it helps you prepare for the DPIA discussion for AI tools straight away.
Four risks an unstructured ChatGPT rollout creates
Rolling out generative AI across an organisation means more than handing out licences. Without a framework, you get a pattern of four predictable problems. The workshop addresses each of them in turn.
Data leaks through consumer versions
Staff paste contract text, customer data or source code into the free ChatGPT.com to get quick help. With the consumer version, input is included for model training by default. Sensitive data thus leaves the organisation without IT noticing. Enterprise licences with no-train guarantees are the structural solution, but only if staff know when to use which version.
Shadow AI without visibility
Marketing uses ChatGPT, sales uses Claude, developers use GitHub Copilot and finance is trying Gemini. All on personal accounts or company credit cards. IT has no overview, cannot enforce single sign-on and does not know who has access to which data. A central tool strategy brings this into the open and consolidates the spend.
Non-compliance with the AI Act
As of February 2025, Article 4 of the AI Act is in force: organisations must ensure AI literacy among everyone who works with AI systems. From August 2026, the rules on general-purpose AI models follow. Without a documented training and governance trail, you will be empty-handed in a regulatory investigation, and high-risk applications call for a more extensive DPIA.
Hallucinations reaching clients
An employee who sends out an AI-generated legal, tax or medical text without review exports the model's reputational risk to your brand. A prompt library with verified templates, plus clear review protocols, prevents unchecked AI output from reaching your clients.
What you have in hand at the end of the working day
The workshop is designed as a working day, not a lecture. Each session has an output that we complete live and that goes home at the end as a draft document. Not an abstract white paper, but working artefacts your organisation can build on tomorrow.
AI governance document
A clear framework setting out who within the organisation makes which AI decisions, how new tools are approved, how we handle DPIAs for AI tools, and which data classification determines which model may be used. Includes an escalation path for borderline cases.
Tool shortlist per use case
A matrix that indicates, for each type of work (writing, analysis, code, customer contact, research, image editing), which tool is best suited, under which licence and for which data classification. Not a one-size-fits-all approach, as ChatGPT, Claude, Copilot and Gemini each have their own strengths.
Training roadmap
A phased rollout of training by staff group: a basic AI literacy module for everyone, an advanced module for power users, and a specialist module for departments that handle sensitive data. With target dates and owners.
Prompt library v1
A first set of verified prompts for recurring tasks in your organisation: email templates, summaries of internal documents, client analyses and reports. Includes instructions on data input and review steps, so that output is reproducible and checkable.
Monitoring protocol
How you measure what AI does in your organisation: usage statistics from the admin dashboards of ChatGPT Enterprise, Microsoft 365 Copilot and Google Workspace, quality signals from the business, and periodic evaluations. Includes a review point to check whether the tool shortlist still holds.
90-day implementation roadmap
Concrete next steps: which licence you purchase in week 2, which training takes place in month 1, when the pilot starts in which department, and which measurement point you schedule for day 90. Nobody leaves the workshop day without a follow-up appointment in their calendar.
Workshop day programme
A typical workshop day lasts seven hours, with four fixed blocks. We work with your own documents, case studies and data classifications, not generic examples, so the output is immediately applicable.
Block 1 — Baseline assessment
Inventory of current AI use by department, existing contracts and licences, data classifications and any previous AI incidents. A shared picture of where your organisation stands today.
Block 2 — Tool selection
Comparison of ChatGPT Enterprise, Claude for Work, Microsoft 365 Copilot and Google Gemini on licensing model, EU data residency, integration options and suitability for each use case.
Block 3 — Governance
Drafting an AI policy, mapping data classifications to tools, and setting up review and escalation protocols. Discussion of the obligations under Article 4 of the EU AI Act and how to document that evidence.
Block 4 — Roadmap
Training roadmap, prompt library v1, a 90-day implementation plan and a monitoring protocol. Recording who carries out which action and when you schedule the first evaluation.
ChatGPT, Claude, Copilot and Gemini: when to use which model
The question is rarely 'which one is the best', because the models are complementary. The question is which combination suits your work processes, data sensitivity and existing IT infrastructure. We discuss these trade-offs during the workshop day, but here are the main points.
ChatGPT Enterprise (OpenAI)
Strong at general text generation, brainstorming, code assistance and multimodal tasks. ChatGPT Enterprise offers no-training guarantees, SAML SSO, admin dashboards and optional EU data residency. Well suited as a broadly deployed tool for knowledge workers; it pairs well with specialist tools such as Copilot for Office work.
Claude for Work (Anthropic)
Stands out for structured reasoning, long documents and careful wording. The 200K context window makes it practical to analyse extensive reports, contracts and codebases. Attractive to legal, audit and analytical departments that value accuracy and reproducibility.
Microsoft 365 Copilot
The most integrated option for organisations on the Microsoft stack. Copilot works directly in Word, Excel, Outlook, Teams and SharePoint and respects existing permissions. There is no separate interface, but AI sits where people already work. It does require a sound information architecture in SharePoint to be truly valuable.
Google Gemini in Workspace
The equivalent for organisations on Google Workspace. Gemini is embedded in Gmail, Docs, Sheets, Meet and Drive and offers strong multimodal capabilities. EU data residency is available within Workspace contracts, and integration with BigQuery and Vertex AI makes scaling up to custom solutions straightforward.
In practice we often see hybrid set-ups: Copilot for everyone on Office, ChatGPT Enterprise or Claude for specific power-user groups, plus department-specific tools for development, marketing or customer service. The workshop day helps you choose that mix deliberately rather than letting it emerge by accident.
Who this workshop is for
The workshop day is aimed at organisations that want to take AI beyond a handful of enthusiastic early adopters. It is not a workshop for a single department but a conversation about the organisation as a whole, which is why the composition of the group matters more than its size.
HR and L&D managers
You are responsible for developing AI literacy among employees and designing training programmes. The workshop gives you a training roadmap, profiles per employee group and the governance needed to make AI learning objectives part of personnel policy.
IT and information managers
You manage licences, data classifications, SSO and information security. The workshop helps you design a tool portfolio that replaces shadow AI with managed alternatives, with clear agreements on data residency, no-train guarantees and monitoring.
Management team and board
You want to know which strategic choices need to be made now so that AI doesn't become a hobby for enthusiastic employees, but a focused productivity drive with measurable returns. The workshop gives you the framework to make that assessment.
Compliance, privacy and legal
You test AI tooling against the GDPR, the AI Act, sector guidelines and internal policies. The workshop gives you the substantive basis to write and maintain an AI policy, plus an initial DPIA outline for the tools you choose.
Operational managers
You notice that AI tools can speed up work processes, but you also know that unregulated use leads to inconsistent quality. The workshop helps you determine, process by process, where AI belongs and which review steps it requires.
Sectors handling sensitive data
Healthcare, finance, legal, government and industry have additional requirements around data residency, audit trails and explainability. We have sector-specific considerations ready for these organisations.
What sets Appfront apart as a facilitator
We are not a pure training organisation; Appfront has been building AI components into software for clients for years. That experience informs every decision we help you make during the workshop, from a realistic picture of what tools can and cannot do to practical integration choices that remain workable beyond the workshop day.
Built, not just advised
Our consultants have sat on both sides of the table: as builders of AI applications for clients and as facilitators of strategic sessions. As a result, the tool shortlist that emerges from the workshop day is not a theoretical list, but a choice you can justify technically.
Model-agnostic advice
We have no exclusive partnership with OpenAI, Anthropic, Microsoft or Google. The advice you receive is based on your situation, not on commission structures. That saves both cost and hassle later on.
Integration expertise for the next phase
Many organisations find that a ChatGPT licence is not enough. Questions arise about integration with your own data, automating repetitive tasks and building internal tools. We have the capacity to help you with that next phase, should you need it.
Privacy, the AI Act and organisational readiness
Introducing AI is not an IT project but an organisational change with legal, ethical and operational dimensions. During the workshop day, we explicitly address the rules that apply to your organisation.
AI Act Article 4: the AI literacy obligation
Since 2 February 2025, the AI Act has required providers and deployers of AI systems to ensure sufficient AI literacy among their staff and any third parties operating the AI on their behalf. The obligation is principle-based, and supervisory authorities expect a documented framework. The workshop produces that framework.
DPIAs for AI tools
A Data Protection Impact Assessment is mandatory for high-risk processing or large-scale use of AI on personal data. During the workshop day, we make an initial inventory of which deployed tools require a DPIA and which parts of it can already be completed.
EU data residency and no-train guarantees
ChatGPT Enterprise, Claude for Work, Microsoft 365 Copilot and Gemini all offer variants with EU data residency and guarantees that customer data will not be used for model training. We discuss which variant per provider suffices for your situation and which contractual terms deserve extra attention.
Organisational readiness
AI rollouts succeed or fail on organisational readiness: clear ownership, a communication plan, a decision model for new tools and a feedback channel for employees. We assess where your organisation is strong and where work remains before you scale up.
After the workshop day: what most organisations choose as a next step
The workshop day is complete in itself: you leave the session with a rollout plan you can carry out yourselves. In practice, we see two common follow-on routes where Appfront can also help.
Strategy and implementation programme
Some organisations choose to have the rollout plan's implementation guided. This includes drafting the final AI policy, preparing DPIAs, setting up governance meetings and managing the first 90 days of the tool transition. We provide senior guidance so that you do not need to create an AI function of your own.
Custom tools and integrations
Other organisations realise during the working day that generic tools don't cover everything. Think of an internal chatbot trained on your own documentation, an agent that processes quotes, or an integration between ChatGPT Enterprise and your CRM. Our development teams build those custom components in follow-on projects.
Frequently asked questions about the ChatGPT workshop for businesses
Ready for a structured ChatGPT rollout?
Book a no-obligation introductory call where we can discuss whether the workshop day suits your situation and which mix of participants will deliver the most value.
Book an introductory call