How to use AI like Claude and ChatGPT at work?

In late 2022, OpenAI launched ChatGPT. Within five days the platform had one million users, faster than any other consumer app in history. Since then, AI assistance has moved well beyond the tech department. It is a daily work instrument that any knowledge worker can pick up and use. The question is no longer whether you will use it, but how to do so smartly and safely.

ChatGPT or Claude: what is the difference?

ChatGPT is developed by OpenAI and is the most widely used AI assistant in the world. The free version runs on the GPT-4o model. The paid version, ChatGPT Plus, adds file analysis, image generation via DALL·E, and live web search. Check current pricing at openai.com/chatgpt/pricing.

Claude is developed by Anthropic. It is known for strong reasoning, careful handling of sensitive topics, and a very large context window, meaning you can feed it very long documents in a single conversation. Check current pricing for Claude Pro at claude.ai.

Both tools are Large Language Models: they generate text based on the instructions you give. They are not connected to your business systems unless you explicitly set that up, and they know nothing about your specific situation unless you tell them in the conversation.

8 concrete ways to use AI at work

1. Writing emails and professional communications

This is the most common use case. Give the AI context: who you are writing to, what the situation is, what you want to achieve. Ask it to draft a message. You review it, adjust it to your voice, and send. What used to take 15 minutes now takes 3. It is especially useful for tricky messages, formal communication in another language, or responses to complaints where tone really matters.

2. Summarising documents and meetings

Paste the text of a long report, contract, or meeting minutes into the chat and ask for a summary of key points, action items, or decisions. Tools like Otter.ai or Microsoft Teams' built-in transcription automatically produce a meeting transcript. Hand that to the AI and get a clean action list in seconds.

3. Drafting quotes, presentations and reports

Describe the context of a project and ask the AI for a structure or first draft for a quote, project report, or presentation. You get a usable skeleton to fill in with the specific details. It speeds up the writing process significantly, even if you end up reworking most of it.

4. Brainstorming and generating ideas

AI is a surprisingly good brainstorming partner. Describe a problem and ask for ten possible approaches, five alternative angles, or a devil's advocate take on your plan. It has not had a bad day, is never tired, and has no personal agenda. That makes it an honest sounding board.

5. Rewriting, simplifying or translating texts

Paste a technical text and ask for a version written for a non-technical audience. Have jargon replaced with plain language. Translate a document from Dutch to French or English. For most European languages, current AI models perform well above Google Translate for contextual and formal texts.

6. Answering questions and explaining concepts

Use AI as a patient tutor for concepts you do not fully understand yet: software tools, regulations, accounting terms, technical jargon. You can keep asking follow-up questions without any embarrassment. That said, for tax, legal, or compliance questions, a human expert always holds final responsibility.

7. Writing or debugging simple code

For employees working with Excel or Google Sheets: ask the AI to write or explain a formula. For anyone working with Python, SQL, or other languages, AI is a powerful aid for writing scripts, tracking down errors, and understanding existing code. This also applies to Power Query M formulas in Power BI.

8. Drafting customer communications and FAQ answers

Give the AI the context of a customer question and ask for a professional, friendly answer that fits your company's tone. Or have it draft a set of FAQ responses based on questions you supply. The result is always a first draft. Fact-checking and final approval stay with you.

What you must NOT do: the GDPR pitfall

This is the most overlooked risk when using public AI tools at work. ChatGPT and Claude, in their free and standard paid versions, are public cloud services. Data you enter may, depending on your privacy settings, be used to train future models.

• Never enter personal data of customers or employees: names, email addresses, financial details
• Do not paste in confidential contracts, business strategies, or pricing structures
• Do not include customer-specific information that is not publicly available
• Always check privacy settings: with ChatGPT you can disable training data use in settings

Want to use AI safely with business data? Enterprise options exist: ChatGPT Enterprise, Microsoft Copilot via Azure OpenAI with data isolation, or self-hosted alternatives where your data never leaves your own infrastructure.

How do you successfully roll out AI in your team?

• Start with volunteers: give interested employees space to experiment and share what they learn with the broader team
• Make it concrete: run a one-hour session where everyone uses AI for their own daily tasks
• Set clear guidelines: what is allowed and what is not, especially around confidential data
• Normalise imperfection: AI output is always a draft, never a finished product
• Measure the time savings: ask employees to track how much time they save on specific tasks

Which tool do you choose for your business?

For individual use without sensitive data, both ChatGPT and Claude work well. ChatGPT has a larger ecosystem of integrations. Claude tends to perform better on longer, more complex texts. Try both: each has a free version and paid tiers — check current pricing on their respective websites.

Want to know how AI can be put to practical use in your specific business context, from customer service to sales follow-up and internal reporting? We are happy to think through an approach that fits your team, your tools, and your goals.