AI is embraced faster than it is understood: how do you consciously use it?

How far apart are usage and knowledge? 

In September, UNESCO and ICOM, the International Council of Museums, published the first global survey on AI usage in museums. Over 400 museums from 90 countries participated. Their conclusion is that museums experiment with AI faster than they create policies, measure results, and train people. 

• 57% of museums use AI, primarily for translation, writing texts, summarizing, and research. 

• 55% have no AI policy or agreements. 

• 17% offer AI training to employees. 

• 65% of AI users believe it is too early to determine AI's impact. 

The gap is not only between usage and policy but also within the sector. Larger museums use AI more often, have more agreements, and train their staff more frequently than smaller museums. Smaller museums mainly cite a lack of time, money, and practical support. You also see that AI is often used by individual employees rather than the organization as a whole. 

This creates an AI divide. Between employees who experiment independently, organizations without a clear direction, and organizations with or without the resources to properly adopt AI.

The same pattern is visible among Dutch citizens. Research by Conclusion shows that many people use AI and think they understand it, while their knowledge falls short. For example, many people don't know that AI can fabricate answers that sound convincing. The report's conclusion: Dutch citizens embrace AI faster than they understand it. Usage grows faster than AI literacy. 

Should you then avoid AI? 

No. UNESCO has been saying the opposite since 2021. In the Recommendation on the Ethics of AI, UNESCO calls on countries to encourage museums, libraries, and archives to use AI. For example, to make collections visible and more accessible to users. Five years later, UNESCO and ICOM continue this approach but with an important addition. After their survey of museums, they write:

In short: using AI is not the goal. The point is to consciously choose when, where, and how AI helps your organization. That requires knowledge. What can AI do and not do? What does it deliver? And what agreements align with your values? That is AI literacy. This knowledge should not only reside with large institutions. Smaller organizations must also be able to participate to prevent the AI divide from growing. 

The museums in the UNESCO survey show what this looks like. They see AI as a tool, not a replacement for expertise. Most ensure a human checks everything before it is shared publicly. A small museum in New Zealand asked its audience what they thought of AI-generated images in promotion. The reactions were mostly negative, so the museum decided against it. That, too, is AI literacy: consciously choosing, sometimes opting not to use it. 

Agreeing together on when to use AI and when not to is a good first step. No AI policy yet? With the guide Create an AI policy for your organization (opens in new tab) and DEN's step-by-step plan, you can get started yourself. 

Get started with AI in your organization 

Do you want your entire team to learn to use AI consciously and safely? DEN offers practical in-company AI training tailored to your organization. Until November 9th, you can get over 70% off.

Not sure what suits you best yet? Schedule a consultation with DEN's AI advisors.