If so, you should be mindful of what information you share with artificial intelligence.

“We should not blindly trust artificial intelligence,” says Professor Kasper Green Larsen.

As AI becomes an increasingly integrated part of everyday life, it also brings new risks. AI can be used for much more than helping with shopping lists, coding, or offering advice. It can also be used for manipulation, fraud, and influence operations on a large scale.

“As humans, we are influenced by what we see online. That means it matters greatly who controls the technology and how it is used,” Kasper explains.

Kasper researches artificial intelligence and the principles behind how these systems are designed and operate. He emphasizes that AI does not develop itself or set its own goals. Instead, it functions according to the patterns and rules that humans have built into it and trained it on.

However, not everyone has good intentions. When AI is developed or exploited by malicious actors, problems such as data breaches, fraud, and manipulation can arise.

You may have received an email from what appeared to be a trustworthy sender asking for your personal information. These are phishing emails, and AI can be used to make them more convincing, tailor messages to specific individuals, and automate parts of an attack. As a result, attackers can operate faster, on a larger scale, and more persuasively than ever before.

“AI is increasingly becoming a target in itself. When someone can manipulate the system or the data it relies on, the technology can also be used as a weapon,” he says.

The professor is also mindful that his own AI research could be misused in the wrong hands.

“Part of my research is focused on reducing the risk of misuse and making AI systems more secure and reliable. For me, the possibility of misuse is not an argument against conducting research, it is an argument for taking responsibility for how the technology is developed,” he concludes.