The proportion of Hungarian social media users who have already tried ChatGPT or the Dall-E 2 image generator is lower than the proportion of respondents in the Czech Republic (25%) and Poland (31%). However, nearly a quarter of respondents aged 18-24 have used AI tools.

The most frequent users are those with higher education (38%), those living in the largest Hungarian cities (40%), and those working in IT (54%), media and advertising (50%), and culture (46%). The oldest respondents, aged 55 and over, very rarely use AI (6%).

The survey clearly shows that interest in this technology declines with age, while concerns about it increase. Some 67% of Hungarians who use social media are concerned that data entered into AI applications could fall into the wrong hands. In particular, people working in public administration (85%), health (74%), education (74%) and finance and banking (72%) point to this risk.

The dangers of artificial intelligence processing users' confidential data are also perceived by 57% of Czechs and 64% of Poles surveyed.

"Cybercriminals are already making millions with poorly crafted fake texts. If you were chatting to an AI bot via instant messaging services like WhatsApp or Microsoft Teams, would you even know you weren't talking to a human? Creating text messages that are deceptively similar to human text messages is just one of the opportunities offered by AI scams," says Chester Wisniewski, CTO at Sophos.

"Stay alert and check all text messages and emails, especially those that ask for login details or relate to financial matters. It's worth taking a few minutes to make sure you know what you're dealing with and if in doubt, ask for help. It's not paranoia – we really are targets for cybercriminals," he adds.

Distinguishing Between Human and AI Content Becoming Harder

About 37% of Hungarian respondents believe that in the next three years they will be able to distinguish between human-generated and AI-generated content. Respondents aged between 18 and 34 (41% on average) and those working in media and IT (51% on average) are the most confident in their judgment. Among respondents aged 55 and over, only one in four are so convinced.

In comparison, only 13% of social media users in the Czech Republic believe they will see a difference between human and AI products in the near future. Among Poles surveyed, the figure is 34%.

One form of artificial intelligence that Hungarians also often encounter is chatbots, used by telemarketing companies, among others.

In a survey by Sophos, 36% of Hungarian respondents said they had clearly recognized that it was a machine at the other end of the line when they were called. This was most often (67%) among respondents aged 18 to 24 and those with a higher education (65%), and least often among older respondents (54%) and those with a lower education (50%).

Almost half of Czech respondents (48%) and 61% of Poles also said they could correctly identify whether they were talking to a person or a machine.

Some 53% of Hungarian respondents believe they will be able to identify who is actually on the phone in the near future, while some 35% said that they were not sure.

The rapid development of artificial intelligence also brings with it the huge potential for misinformation. Those most susceptible to fake content generated by AI are those who claim to be able to distinguish bots from humans without any particular problem, Sophos notes.