ChatGPT at work: Crafting an effective Company Policy for AI usage

This article was updated January 2025

ChatGPT at work: Revolutionising Today’s Digital Age

The integration of artificial intelligence (AI) technologies has become increasingly common in the workplace. While AI brings numerous benefits to organisations, it also raises concerns regarding the appropriate use of AI systems and the potential misuse of confidential information.

Consider the latest advancements in workplace AI tools, such as Microsoft Copilot, Google Gemini, and ChatGPT’s latest versions (e.g., GPT-5). Copilot is now embedded across Microsoft 365 apps like Word, Excel, PowerPoint, Outlook, and Teams, offering enhanced features such as real-time summarisation, AI-driven insights, and collaboration suggestions. It includes Business Chat, which seamlessly integrates across your calendar, emails, chats, documents, meetings, and contacts to streamline workflows.

Establishing a comprehensive company policy that addresses the use of ChatGPT and AI at work remains crucial to ensure ethical behaviour, protect sensitive data, and maintain a productive work environment. Outlined below are key considerations for drafting a workplace policy, including examples of confidential information to avoid using in ChatGPT and the potential consequences of misusing such information.

1. Positive Ways to Use ChatGPT and AI Tools at Work

Training and Onboarding:

ChatGPT can support new employees during the onboarding process by answering their questions, providing training materials, and offering guidance. It can offer personalised learning recommendations and assist in knowledge transfer.

Content Generation:

ChatGPT can be used to generate content for various purposes, such as writing blog posts, drafting emails, creating social media updates, or developing marketing materials. It can help with brainstorming ideas, proofreading, and improving the overall quality of written content. Additionally, many organisations now leverage AI to create dynamic templates for internal communications and external outreach.

Language Translation:

If your workplace deals with international clients or colleagues, ChatGPT can assist in real-time language translation. It can help overcome language barriers and enable effective communication across different languages. For example, tools like DeepL and ChatGPT’s multilingual features now provide highly accurate translations tailored to professional settings.

Enhancing Customer Service:

ChatGPT and other AI systems can serve as a virtual assistant to handle customer queries, providing real-time responses and assistance. It can handle frequently asked questions, provide troubleshooting tips, and offer personalised recommendations. More advanced tools now integrate seamlessly with CRM platforms to enhance customer experiences further.

Automating Repetitive Tasks:

AI can be employed to automate routine and time-consuming tasks, allowing employees to focus on more value-added activities. This boosts productivity, efficiency, and employee satisfaction. For example, integrating AI with marketing tools like make.com enables automated tracking and CRM updates from ad campaigns.

Data Analysis and Insights:

AI technologies can analyse vast amounts of data, providing valuable insights for decision-making, identifying patterns, predicting trends, and driving innovation. Generative AI can now visualise data in real time, creating tailored reports and dashboards for quicker decision-making.

Workflow Optimisation:

AI can optimise workflows by streamlining processes, identifying bottlenecks, and suggesting improvements, thereby improving overall operational efficiency.

2. Confidential Information to Avoid sharing with AI

Personal Identifiable Information (PII):

Employees should refrain from using PII, including tax ID’s, addresses, phone numbers, or any other personally identifiable information of individuals. Unauthorised use or disclosure of PII can lead to severe privacy breaches and sometimes has legal consequences. Many companies now implement AI firewalls or restrict the use of external AI tools to prevent PII leakage.

Intellectual Property (IP):

Confidential company information, trade secrets, copyrighted material, or any proprietary data should not be shared with ChatGPT or any other AI system. Protecting intellectual property is crucial for maintaining a competitive advantage and safeguarding business interests. AI providers like OpenAI now offer options to disable data sharing, but companies should still implement internal safeguards.

Financial Data:

Sensitive financial information, such as bank account numbers, credit card details, or undisclosed financial reports, must not be shared with ChatGPT. Unauthorised disclosure of financial data can result in financial losses and damage to the organisation’s reputation.

Health Information:

Medical records, patient data, or any other personally identifiable health information should not be utilised in ChatGPT. Breaches of health information can lead to legal consequences and are a violation of privacy laws.

3. Harmful Ways to Use AI at Work

Unfair Bias:

If AI systems are not properly designed and tested, they may inadvertently perpetuate biases present in training data, leading to unfair treatment of employees, customers, or stakeholders. This can result in discrimination claims and damage to the organisation’s reputation. AI audits are now increasingly used to identify and mitigate bias in systems.

Invasion of Privacy:

AI technologies should not be used to intrude on employee privacy rights, such as unauthorised monitoring of communications, personal devices, or sensitive personal information.

Unethical Data Collection:

Using AI at work should not involve unethical data collection practices. This includes obtaining personal information without consent, collecting excessive or irrelevant data, or leveraging AI to exploit user privacy for unauthorised purposes. Such practices not only violate privacy regulations but can also lead to distrust among employees and customers, resulting in reputational damage and potential legal repercussions.

Conclusion:

When implementing ChatGPT and AI at work, having a well-defined company policy is essential to guide employees in the appropriate use of these technologies. The policy should clearly outline the confidential information to be avoided while using ChatGPT, emphasising the potential consequences of misusing such information. Additionally, the policy should highlight the positive ways in which AI can enhance productivity, customer service, data analysis, and workflow optimisation. By establishing ethical guidelines and fostering responsible AI usage, organisations can harness the benefits of AI while safeguarding privacy, protecting sensitive information, and promoting a positive work environment.

The evolving landscape of AI tools and regulations makes it even more critical for organisations to stay informed and proactive in their AI adoption strategies.