The use of AI is increasing in Sweden, but at the same time, three out of four have no or very low confidence that AI services handle personal information securely. This is shown in the report "The Swedish People and AI 2025".
39 percent have at some point asked a chatbot for personal advice, most commonly about illness. At the same time, 54 percent say they generally don't trust the information AI provides to be true.
We have also had this duality with social media, yet we cannot help but use them, says Åsa Wikforss.
More dangerous than social media
But in one respect, Wikforss describes chatbots as “extra seductive” and more dangerous than social media:
"They enter into a dialogue with us. It becomes personalized, they take it a step further than social media," she says, continuing:
In a way, it's antisocial to have a dialogue with a machine. But it gives the impression that we're having a dialogue with a human. It's hard to resist.
Wikforss does not see it as surprising that many in the report express skepticism about AI but still use the services.
We've seen that with social media too. People's self-esteem is quite poor. For example, we know that young people think they are good at distinguishing between fake content and real content, but that doesn't match their actual skills.
“A tool”
Wikforss points out that AI in itself does not have to pose any problems in people's everyday lives. Rather, it is about how the tools are used.
We must not lose our own thinking. See it as a tool. If you have knowledge of how to use the services, they can be helpful and get you to the right place.
But we should worry about putting our intelligence to work for someone else to think, calculate, read and analyze for us. We need to use our own brains for that.
The report “The Swedish people and AI 2025” is based on responses produced by Verian via the Sifopanelen. It is based on 1,000 interview responses from people between the ages of 16 and 70 in Sweden and was conducted in October.
The survey was produced by Insight Intelligence, SKR, Esam, Amazon Web Services and Bonus Copyright Access.
The report states, among other things, that:
+ 22 percent say they use AI every day. 34 percent say they use AI at least once a week. Up from 14 and 22 percent respectively in 2024.
+ 39 percent have asked AI for personal advice.
+ 74 percent have no or fairly low confidence that AI services handle personal information securely.
+ 54 percent do not trust at all or quite a bit the information that AI provides.
+ 35 percent are comfortable with AI being used as support in healthcare.
Source: Insight Intelligence
