RISJ releases new report on how people are using AI

The Reuters Institute for the Study of Journalism (RISJ) has published a report on the results of recent AI research. 

Written by Dr Richard Fletcher and Professor Rasmus Kleis Nielsen, the report is based on the results of an online survey which examines how the public uses – and what it thinks of – AI. 

The survey sought public opinion in six countries: Argentina, Denmark, France, Japan, the UK, and the USA.

Through the online survey, the RISJ aimed to discover:

• to what extent the public uses generative AI
• how people use AI
• what the public think of the use of AI in journalism and other areas of work and life

In this article, we describe some of the key points from the report.

To what extent does the public use generative AI?

Given its innumerable mentions in the media, it comes as no surprise to find that ChatGPT is by far the most widely recognised generative AI product. In fact, 50% of the people surveyed in the six countries have heard of it. 

However, only 2% of the population use ChatGPT frequently in the UK and France. The product is used least in Japan, where only 1% use it routinely. However, it is most popular in the USA, where 7% of those surveyed use the product frequently.

Despite these seemingly small numbers, ChatGPT is two or three times more popular with users than the next most widely used products, Google Gemini and Microsoft Copilot. 

More popular with younger people

56% of people aged between 18 and 24 say they’ve used ChatGPT at least once, whereas only 16% of those aged 55 and over have used it.

How do people use AI?

In all of the six countries, roughly 28% of people said they use generative AI to create different kinds of media – mostly text, but also including code, images, video and audio.

24% of respondents across the six countries said they use these tools for getting information, whereas just 5% use them to get the latest news. 

What does the public think of AI?

The survey results show that most people expect AI to have an enormous impact on virtually every aspect of society over the next five years. Many people believe that AI will have a substantial impact on:

• political parties (51%)
• news media (66%)
• science (66%)
   

How do people think AI will impact our lives?

Opinions vary on whether AI will be used responsibly within different sectors in future. Roughly 50% of those surveyed expect scientists and healthcare professionals to use it responsibly, whereas less than 33% trust politicians, social media companies and news media to do the same. 

For the most part, there appears to be very little difference in how people of different ages, genders, and education levels, foresee the impact of generative AI on ordinary people. However, younger respondents are more likely to expect a direct impact on their own lives than older people expect on theirs.

In addition, the survey reveals a negative percentage point difference (-18) between those who are optimistic about the impact of AI on job security and those who are pessimistic about it – indicating that more people believe the technology will ultimately have a detrimental impact on job security.

Do people trust AI’s output?

There are several documented cases of generative AI giving inaccurate answers in response to factual questions. With this in mind, it is perhaps somewhat fitting that users tend to say they ‘somewhat trust’ AI’s answers – indicating a degree of scepticism. 

AI and journalism

How does the public think AI is affecting journalism right now?

The RISJ survey asked respondents how often they believe journalists use AI with ‘some human oversight’. 

The results showed that 43% believe that journalists often use AI ‘to edit the spelling and grammar of articles’, while 42% of people say they believe journalists sometimes use AI for ‘creating an artificial presenter or author’. Meanwhile, 43% of people said they believe journalists always use AI for translation.