10 rules for handling AI in business

Successful strategic interaction with artificial intelligence (AI) requires much more than just the ability to use tools. A clear stance, defined processes and specific interaction techniques are important. Here are ten key points your company should consider in order to be strategically successful when dealing with generative AI.

Christian Hansen

10/4/20252 min read

10 rules for handling generative AI in the company.

1. Define a clear strategy and leadership role

The implementation of generative AI affects the core communication function and potentially the entire value chain. Senior management must shape and steer AI development. It is crucial to establish a clearly defined AI strategy at management level that encompasses the vision, goals, areas of application and risks. Your communications department should set an example within the company and lead the way in implementation along the entire value chain.

2. Promote systematic upskilling and adaptive culture

Generative AI is changing the skills profile of employees. To strengthen digital resilience, you need systematic upskilling and the targeted promotion of an adaptive culture. Skills such as critical and holistic thinking, technological know-how, an understanding of data and prompting abilities are essential for employees.

3. Strategically utilising gains in efficiency

AI enables increases in efficiency and productivity. However, the resources gained as a result should not simply be saved away. Actively use this freed-up time for strategy development, consulting and personal stakeholder management.

4. Establish responsible AI governance and risk management

The use of AI carries significant risks, including data protection, ethics and factuality. It is essential that you develop clear guidelines and training for using AI tools in order to manage these risks. Embed principles such as fairness, transparency and accountability in your AI governance and review processes for structural injustices before AI replicates them.

5. provide precise and specific prompts

The quality of the output depends directly on the clarity of the input (‘Shit in, shit out’). Prompts should be concise, clear and simple, avoiding unnecessarily complex language. Be very specific about the instruction and the task you want the model to perform. Avoid vague descriptions; direct communication is more effective.

6. Clearly define the context, role and output format

Assign the AI a specific role (persona), e.g. as an expert in your industry. Provide the necessary context and target audience (e.g. beginner or professor) to obtain relevant and in-depth answers. Explicitly specify the desired output format (e.g. list, table, step-by-step instructions) to control the structure of the results.

7. Use iteration and continuous feedback

Prompting is an iterative process. Start with simple instructions and then gradually add details to achieve better results. Critically evaluate the output, give the AI feedback (e.g. if the answer is too general) and adjust the prompt.

8. Apply advanced thinking methods

Use techniques that encourage the model to reason. For example, through ‘chain-of-thought prompting’ (‘Let's think step by step’). This increases the transparency of the argumentation and improves performance on complex tasks. For very complex workflows, you should break the task down into several sequential prompts (prompt chaining) to maintain context and build on previous results.

9. Actively challenge AI and provoke dialogue

Consider AI as an interactive mirror and a ‘reflection monster,’ not just a pure information tool. Consciously challenge AI to provide counterarguments, examine your ideas from a critic's perspective, or simulate tactical objections from the target audience. This helps to uncover blind spots and gaps in your argumentation.

10. Integrate your company's tone of voice

Use generative AI to implement and align tone of voice and core messages across the company. By bringing together topics, messages, language and imagery in a company-wide communication platform, integrated communication can be taken to a whole new level. Prioritise business solutions to protect company data and better align communication with your brand worlds.

If you need assistance in using generative AI more successfully for yourself and your company, please feel free to contact us. We provide pragmatic, goal-oriented help without complicated buzzword bingo. Get in touch!