How can we raise awareness among teams about AI risks?​

This isn’t an exception specific to your company—no… In every company, the use of AI is spreading faster than the frameworks to govern it. On a daily basis, one person (discreetly) pastes a contract excerpt into a chatbot to summarize it, another (distractedly) uses a figure without verifying its source, or (lazily) uses a generator for a task that could have been done differently. It’s often in these ordinary, seemingly innocuous actions that the risks associated with AI lurk. Translated with DeepL.com (free version)

Data privacy, errors and hallucinations, biases, environmental impact: To raise teams’ awareness of AI risks, we must provide them with guidelines they can apply in their work, based on situations they are likely to encounter in reality.

Start with real-world usage to make the risks tangible

However, not all jobs face the same risks. For example, a sales team does not handle the same information as an HR or legal department, just as failing to verify a response intended to prepare for an internal meeting does not raise the same concerns as including a figure in a sustainability report.

We therefore believe it is most useful to encourage employees to reflect on these situations: What information can be shared with an AI? How can we verify an answer that seems perfectly credible—except to someone who knows it isn’t? In what cases is it necessary to maintain human oversight—or, better yet, human intervention? And above all, the first question to ask is: Is the use of AI truly relevant to the task at hand?

By comparing their practices and perspectives, teams gain a better understanding of the risks and learn to identify the right reflexes, which in turn makes it easier to adopt common guidelines for use.

Making Awareness-Raising a Collective Experience

To that end, we have created a new scenario that can be adapted to any team and any organization: iAïe! In just a few hours, participants can examine their own practices and delve into the specifics of how they use technology to understand the specific challenges involved and develop solutions to better manage the risks they had been taking without realizing it. And the most paradoxical (yet effective) part of it all? They do this by drawing on the expertise of… AI characters!

It is this unique feature that makes the experience truly productive. On the one hand, participants can interact freely with their virtual teammates to ask as many questions as they like in order to fully grasp the subject. On the other hand, this freedom comes with consequences: they will be able to see, in real time and with complete transparency, the resource consumption resulting from their conversations.

This is an opportunity to realize that these uses are by no means trivial, even though eSqwad’s AI is designed to be particularly responsible: our AI models are hosted and run exclusively in Switzerland (outside any extraterritorial jurisdiction and outside the scope of the U.S. Cloud Act) to ensure absolute sovereignty and confidentiality; the conversations, questions, and documents analyzed are never used to train the AI models, nor are they retained by the provider. Furthermore, our AI is hosted in a data center powered by 100% certified renewable energy, operating without artificial air conditioning or water evaporation, and uses an optimized European open-weights model, in which 100% of the heat generated by the servers is recovered and reused to heat local homes.

Combining rigor and fun: the solution offered by eSqwad.

This approach allows us to address AI risks in real-world scenarios, by highlighting the issues and trade-offs that employees will also need to manage in their day-to-day work. It also provides an opportunity to collectively discuss the limits to be set, the level of oversight required, and the role we wish to assign to AI in professional practices.

Awareness-raising can thus directly inform the company’s usage policy, because the rules make more sense when they are linked to situations that the teams have had the opportunity to explore together.

Raising teams’ awareness of AI risks means empowering them to make informed choices in their day-to-day use of the technology. This is precisely what a collective experience centered on the company’s challenges enables, and it is eSqwad’s mission: to put artificial intelligence at the service of collective intelligence, through the thoughtful use of this technology.

To learn more about this topic, check out our article on Risks of AI in the Workplace and Technology Choices Related to CSR.