Some people recognise the problems that AI can create within an organisation before others do. They see colleagues entering information into tools without checking what happens to that data, relying on AI-generated answers without verifying them, or adopting tools without anyone assessing the risks. They know something needs to change, but they do not hold a position that allows them to drive that change.
This situation is more common than it might seem. You can have sound judgment, experience and a clear vision of what needs to be done, but that does not mean you can impose it on the rest of the team.
So, how can you help your organisation use AI more responsibly when you have no authority to make decisions for others?
The influence that does not appear on the organisational chart
In every organisation, there are people who influence how their colleagues work without holding management positions. Their influence comes from trust, consistency and their ability to offer useful solutions.
When you use AI, for example, you can explain why you verify its responses, what information you choose not to enter into a tool, or when you believe it adds too little value to justify its use. By doing this naturally and sharing your reasoning when appropriate, you help others consider questions they may not have thought about before.
This is also a form of leadership. You do not need to have all the answers or become the person who warns everyone about the dangers of technology. You simply need to apply sound judgment, explain your decisions and be willing to share what you have learned.
However, individual example has its limits. It can open conversations and help change habits, but decisions about tools, data, training and responsibilities need organisational support if those changes are to become established.
Why leading by example can change habits
An instruction can get someone to follow a rule while they feel they are being watched. Changing a habit requires people to understand why doing things differently is worthwhile.
Imagine a colleague using an AI tool to prepare a report and accepting every answer at face value. You could simply point out that they should check the information, or you could share how you review the output, what mistakes you have encountered and what steps you take before using the results.
The second approach offers something the first does not always achieve: a concrete way of working.
The same applies to privacy. Before entering information into a tool, you can check what data it contains, what terms of use apply and whether the tool is approved for that task. This shows that responsible AI use is part of everyday professional practice.
Leading by example does not guarantee that others will change, but it can encourage them to start asking questions. Those questions are a good starting point for improving how a team works.
How to get started when you are not in charge
The first step is to identify a specific situation in which AI use could be improved. Trying to change an entire organisation’s relationship with technology overnight often leads to frustration and scattered efforts.
It could be a tool that the team uses without understanding its terms, information that is being shared without sufficient care, or AI-generated output that is added to documents without proper review. The more specific the situation, the easier it is to explain why it deserves attention.
Next, focus on your own practice. Use tools thoughtfully, verify their output and follow data protection requirements and internal procedures. If you identify a risk, explain what you have observed and suggest an alternative instead of simply pointing out the problem.
It also helps to find people who share your concerns. Someone in IT, human resources, the legal department or another team may already be thinking about similar issues. A conversation between several people can give a proposal more momentum and help identify who is responsible for making the decision.
The goal is to turn concern into a practical proposal. If a tool raises questions, you can suggest reviewing it. If the team lacks the knowledge to use it properly, you can propose a training session. If nobody knows which tools are approved, you can suggest clarifying the procedure.
One important point to remember is that identifying a risk does not automatically make you responsible for resolving it. If you spot a potential security, privacy or compliance issue, raise it through the appropriate channels with the people who have the authority to act.
The risk of trying to convince everyone
When you identify a problem before others do, it is easy to become frustrated when they do not react with the same urgency. You may find yourself repeating the same warnings, interpreting the lack of response as indifference, or feeling that you are the only person who cares about the issue.
However, every team has its own priorities, knowledge and constraints. Someone may use a tool inappropriately because they do not understand its limitations, have not received training or are under so much pressure that they focus on finding the quickest solution.
Understanding these circumstances does not mean downplaying the risks. It means finding a more effective way to address them.
Rather than trying to convince everyone at once, focus on the problems with the clearest consequences and the people who are willing to listen. Explaining how a particular practice can affect someone’s work, customer data or the quality of a decision is often more useful than starting a general debate about the dangers of AI.
Patience also has its limits. Some situations require a formal decision, particularly when personal data, confidential information or decisions affecting other people are involved. In these cases, informal influence can help initiate change, but it cannot replace established procedures or the organisation’s responsibilities.
What it takes to build a culture of responsible AI use
A culture of responsible AI use develops through everyday decisions, but it also requires the right conditions to sustain those decisions over time.
Organisations need to clarify which tools employees can use, what information they can enter, who can approve new uses and when human review is required. They must also provide training so that people understand the capabilities and limitations of these technologies and know how to apply them to their work.
People who do not hold management positions can contribute significantly to this process. They can identify needs, share good practices, spot problems and put forward proposals to the people responsible. They can even help shift the conversation about AI away from tools alone and towards the decisions made with them.
However, the responsibility for establishing the right conditions belongs to the organisation and those with the authority to make decisions. It should not depend on one particularly committed person making endless efforts to compensate for what is missing.
Individual judgment can start a conversation. Training helps turn it into a shared practice, and governance gives that practice continuity.
Frequently asked questions
How can I promote responsible AI use in my organisation if I am not a manager?
Start by identifying a specific situation that could be improved, apply good practices in your own work and share your reasoning with the team. If you identify a problem, propose a practical solution and raise issues requiring a formal decision with the people responsible.
How can I discuss AI risks without sounding anti-technology?
Focus on concrete situations and explain what can be done to manage them. You can show how to verify an answer, protect confidential information or choose the right tool for a task. The goal is to help people use technology thoughtfully, not to make them reject it.
What can I do if management ignores my proposals?
Present the problem with examples, explain its consequences and suggest a specific course of action. If possible, find other people who share your concerns and can contribute their perspective. If the issue involves security, privacy or regulatory compliance, use the established internal channels to raise it. A lack of response does not make you responsible for something that belongs to the organisation.
Can one person change an organisation’s AI culture?
One person can initiate change, influence colleagues and help good practices spread. Sustaining those changes requires time, training, clear procedures and organisational support. Individual leadership can open the way, but the organisation must play its part if the change is to last.
