Google Trends shows that searches for terms like “ChatGPT” and “AI” are at all-time highs, according to a recent report from DISQO, but do humans really trust artificial intelligence — specifically generative AI technology?
Not so much. Trust is pretty bad across every consumer group participating in a DISQO survey of 28,376 members of the community conducted between February 28, and March 2, 2023.
There’s some room for optimism. Trust in AI and human-generated content is about the same for 25% to 35% of consumers. That leaves nearly 60% of individuals who trust AI-generated content less, with more than half of those saying they trust it much less.
For brands seeking to expand their footprint in AI-generated technology, it follows from the 28,376 data that trust is easy to lose and extremely hard to win.
To analyze the perceptions of generative AI, DISQO gathered input from consumers across four areas, surveying28,376 members between February 28, and March 2, 2023.
Data was weighted to a representative U.S. audience on age, gender, and income. Those areas include general awareness and use, perception by content area, trust and distrust, and disclosure expectations.
The study began by asking about their overall knowledge of, use, and exposure to generative AI tools like ChatGPT.
Overall, 6% were highly familiar with the technology, 26% were comfortable with their knowledge, and 68% said they had low knowledge of the technology. Even when broken down into generations and education, less than 10% of consumers state that they have high knowledge of these tools, and that number drops below 5% in certain audiences such as older adults, women, and high school educated.
Males are more familiar than females with the technology. Females have less desire to use tools like ChatGPT to generate AI-written content, compared with men.
Similar to knowledge, use has been limited. Some 10% of consumers have used generative AI tools. This rises as high as 20% among Gen Z. Among early adopters, overall consumer experience reactions have been good, with about 70% having a positive experience with the technology.
Most people participating in the survey are unsure about whether they have seen the technology or have read AI-generated content. Only about 25% of consumers believe that they’ve read media content that was generated by AI tools.
The data suggests that it likely underestimating how much content is influenced by generative AI tools. It also reflects the relative “ignorance” of what marketers and brands do and a “naive optimism about the current or future state of content marketing, as a large swath of content will eventually be driven by advanced AI technologies,” per the study.
It turns out many consumers are not open to AI-generated content tools. Some 34% don’t think generative AI tools should be used for many of the marketing services marketers talk about today.
Where is AI-generated content appropriate? Of the top five, 25% said marketing materials, 18% cited online courses, 17% said corporate emails, 14% said social media posts, and 13% cited news articles.
Older generations such as Baby Boomer are much less open to AI being used across these areas - 43% are not open to this technology in any of these areas, which is 21 points higher than Gen Zers. That topline observation reinforces some of the age-related conclusions we’ve seen elsewhere. Not all areas show a negative trend by age.
What are your biggest concerns about content that’s generated by an AI system? Some 45% were concerned about poorer accuracy, 38% about lack of human touch, 36% about negative impact on jobs, 35% on low emotional depth, 29% on more bias, 27% pointed to lack of creativity, 10% cited lack of humor, and 11% were not concerned at all.