In the past few months, AI language tools have taken the world by storm. The likes of ChatGPT, launched in November last year, and Google Bard, launched this March, have been used by everyone. From University students asking AI to write their academic essays, to email marketing platforms using AI to craft smart, undetectable marketing communications.
But is AI-generated marketing actually undetectable? To find out, we surveyed 510 Brits to see if they could spot the difference between brand marketing emails written by AI and those written by an actual human copywriter.
We recruited ChatGPT to churn out three emails for a made up sustainable beauty brand. Each email had a different purpose: there was a promotion/sale announcement, a new product launch and a loyalty scheme reminder. We then asked a copywriter to do the same.
Respondents were asked to guess whether each of the six emails were generated by AI or a human, and the results were interesting.
1 in 2 people CAN’T Spot a Fake Email
Of those who responded to the survey, less than 2% (1.18%) were able to guess all six emails correctly, suggesting that the majority of consumers aren’t able to tell the difference between AI generated marketing communications and those created by a human.
In fact, respondents were wrong about emails half the time, with more incorrectly guessing that the copywriter’s emails were AI (52%) than vice versa (48%). Overall, one in two people were unable to guess the difference between AI and human-generated marketing copy. Interestingly, more men (51%) than women (48%) were able to guess correctly.
Gen Z is more likely to spot an AI generated email
Those aged 18 to 24 (which made up 14% of respondents) were much better than other age groups at spotting AI-generated emails, guessing correctly more than half the time (53.7%). Clearly, generation Z is more acutely aware of the differences between AI copy, and copy written by a human. That said, 18 to 24-year-olds were still wrong more than 46% of the time.
Unsurprisingly, those aged 55 to 64 were the worst at spotting AI-generated emails, but only marginally, with the majority (52%) still guessing correctly. Some 48% of this age group were unable to tell the difference.
Age group results:
| Age Group | % of Correct Responses | % of Incorrect Responses |
| 18-24 | 54% | 46% |
| 25-34 | 49% | 51% |
| 35-44 | 50% | 50% |
| 45-54 | 48% | 52% |
| 55-64 | 48% | 52% |
| 65+ | 51% | 49% |
People living in the North East were the worst at spotting AI generated emails
Those who live in the North East found it the most difficult to spot AI-generated emails, with people in this region guessing incorrectly the majority of the time (58.3%). Northern Ireland is home to the most AI-savvy people, with those living in NI guessing correctly more than half the time (57%).
Yorkshire and The Humber and Wales also fared quite badly, with people from these regions guessing incorrectly more than half the time (55% and 52% respectively).
Regional results:
| Region | % of Correct Responses | % of Incorrect Responses |
| North East | 42% | 58% |
| Yorkshire & The Humber | 45% | 55% |
| Wales | 48% | 52% |
| London | 49% | 51% |
| South East | 50% | 50% |
| East Midlands | 50% | 50% |
| Scotland | 50% | 50% |
| North West | 50% | 50% |
| South West | 50% | 50% |
| East of England | 51% | 49% |
| West Midlands | 53% | 47% |
| Northern Ireland | 58% | 42% |
Only 26% show concern for AI use in brand comms
When asked if they were unhappy about the idea of AI-generated marketing and communication campaigns, the majority of people (73%) were unphased. Those aged 55 to 64 were the least likely to care about AI being used in advertising, with just 20% saying they were unhappy about it.
Only a quarter of respondents said they would mind. Those aged over 60 were more likely than those aged 25 to 64 to feel unhappy about the use of AI in brand communications, with 30% answering “yes”.
Interestingly, it was actually the younger Gen Z respondents, aged 18 to 24, who had the biggest problem with receiving AI-generated brand communications, with nearly 40% voting “yes” (37.5%), which reflects the notion that Gen Z wants authenticity from brands and businesses.
What does this mean for the future of email marketing?
This experiment has real implications for the future of digital marketing. Clearly, the public is torn, with the results being completely split. That said, it’s clear the distinction between AI-generated content and content created by humans is becoming more and more indiscernible, and less and less people care either way, which makes AI a useful addition to any email marketers arsenal.
Younger generations are more savvy when it comes to understanding the difference between AI communications and marketing written by humans. This could mean that email marketing will need to keep an element of authenticity if it is to appeal to the audience of the future. Without the emotion, empathy and personalisation that AI simply can’t deliver, email marketing would lose its human touch.
Summary
Language processing AI tools like ChatGPT and Bard have come a long way already, and it’s easy to see how they’ve been able to infiltrate many occupations, not least copywriting. A small-scale survey found that the majority of people can’t always tell the difference between a marketing email generated by AI and one written by a human, although this differs slightly by age and gender. Younger generations are more likely to spot an AI-generated email, and they’re also more likely to be unhappy about AI taking over brand communications. Although over-60s were some of the worst at telling the difference, this group also had reservations about AI-generated marketing emails.
Overall, the majority of people can’t tell the difference between an AI-generated marketing email and one written by a human, and the vast majority (74%) don’t mind if their communications are generated by AI.
Want to see how you fare? Take the quiz here.
Methodology
A survey of 510 participants from across the UK, varying in age and gender, were shown six marketing emails for a fake sustainable beauty brand. Three of the emails were generated by ChatGPT and the other three were written by a copywriter. Both had the same brief.
On a digital survey, participants were asked to select whether each email was written by an AI or a human.
Participants were also asked whether they minded if brand communications were generated by AI instead of humans.