Artificial intelligence Study: Distinguishing between humans and AI is difficult

Source: dpa 1 min Reading Time

Related Vendors

Does a machine or a person answer? The increasing performance of systems with artificial intelligence is causing many people in Germany to worry that they will no longer be able to recognize them accurately. This is a key finding of a representative opinion poll conducted by the Allensbach Institute for Public Opinion Research on behalf of Deutsche Telekom (German telecommunications company).

Chatbots pass the Turing test and can hardly be distinguished from "real" people.(Image: sdecoret—stock.adobe.com)
Chatbots pass the Turing test and can hardly be distinguished from "real" people.
(Image: sdecoret—stock.adobe.com)

According to the survey, around three quarters of people in Germany aged 16 and over have already heard of AI programs such as ChatGPT or Google Gemini. In this group, two thirds of respondents believe that it has become difficult to distinguish between a human communication partner and an AI chatbot.

Man or machine? Hard to say

The majority of those who have already used AI programs also find it difficult to distinguish between humans and machines. 57% of users of such chatbots agree with this statement. Among regular users, the fear of no longer being able to distinguish between a human conversation partner and a software robot is slightly less pronounced.

Every second user who uses an AI system at least once a week shares this fear. And one in five regular users of AI chatbots have actually sometimes forgotten that they are communicating with a machine when communicating with such programs.

The study is based on a representative survey for which 1,040 people over the age of 16 across Germany were questioned verbally in face-to-face interviews in September 2024.

Turing test passed

If chatbots with AI behave in such a way that they can hardly be distinguished from humans, then they have passed the so-called Turing test. The British computer scientist Alan Turing developed a corresponding test scenario back in 1950. According to this, artificial intelligence is always present when a questioner can no longer distinguish whether a human or a machine is formulating the answers. At the time, Turing predicted that this would be the case around the year 2000.

Subscribe to the newsletter now

Don't Miss out on Our Best Content

By clicking on „Subscribe to Newsletter“ I agree to the processing and use of my data according to the consent form (please expand for details) and accept the Terms of Use. For more information, please see our Privacy Policy. The consent declaration relates, among other things, to the sending of editorial newsletters by email and to data matching for marketing purposes with selected advertising partners (e.g., LinkedIn, Google, Meta)

Unfold for details of your consent