1. Use AI for General Knowledge and Brainstorming
The safest way to use public AI tools like ChatGPT or Google Gemini is to treat them as a creative partner for non-sensitive tasks. Use them to brainstorm marketing slogans, draft a generic social media post, or explore topics for a presentation. The key
is to ensure your prompts contain no proprietary information. For example, instead of asking, “How can our new widget XYZ beat our competitor's model ABC?” ask, “What are some innovative marketing strategies for a new tech gadget?” This approach leverages the AI's vast knowledge base for ideation and creative problem-solving without ever exposing internal strategy or product details. Think of it as a tool to sharpen your own thinking and overcome creative blocks on public-facing concepts.
2. Summarise Publicly Available Information
Generative AI is incredibly effective at summarising long documents, articles, and reports. This is a low-risk activity as long as the source material is already in the public domain. For instance, you can ask an AI tool to summarise a lengthy market research report published online, a new government regulation, or a collection of news articles about an industry trend. This saves you valuable reading time and helps you quickly grasp key points. The critical rule is to avoid pasting the text of private documents. Instead, if the tool has web-access capabilities, provide a public URL. This method keeps your internal documents, such as client emails or unpublished reports, completely out of the system.
3. Leverage AI for Skill Development
You can use AI as a personal tutor to learn new skills that benefit your role. Ask an AI to explain complex concepts, such as a statistical method or a marketing framework. Programmers can use it to debug generic code snippets or learn a new coding language. For this to be safe, the code or concepts you input must not be proprietary. For example, pasting a unique algorithm from your company's software is a major risk. However, asking, “Explain how a ‘for loop’ works in Python with an example of iterating through a list of numbers,” is perfectly safe. This approach turns AI into a powerful, on-demand learning and development tool that accelerates your professional growth without compromising company secrets.
4. Anonymise Data Before Inputting It
In some cases, you may need AI to analyse a dataset that contains sensitive information. The only way to do this with a public tool is to meticulously anonymise the data first. This means removing all personally identifiable information (PII) like names, addresses, and phone numbers, as well as any other context-specific details that could be used to identify a person or company. For instance, change “John Doe from Acme Corp in Mumbai” to “a user from a large company in a major city.” However, this method carries a significant risk if not done perfectly, as AI can sometimes re-identify data from seemingly innocuous details. This should be used with extreme caution and is generally less preferable than using a secure, company-approved tool.
5. Use Approved, Enterprise-Grade AI Tools
The most robust and secure way to use AI with company data is to use an enterprise-grade platform that your employer has vetted and approved. Services like Microsoft Copilot for Microsoft 365, ChatGPT Enterprise, and Google Workspace AI are specifically designed for business use. These platforms typically include contractual guarantees that your company’s data will not be used to train their public models and will remain private. These tools often integrate directly and securely with your company's documents and systems, allowing you to ask questions about specific projects or analyse internal data with a much lower risk of leaks. Always check your company’s IT policy to see which, if any, AI tools are approved for use.
