Survey: 88% find it harder to tell real from AI content

Malwarebytes surveyed 1,500 adults aged 18 and older across the United States, United Kingdom, Austria, Germany and Switzerland and found widespread doubt about online content. The sample was balanced for gender, age and region. Eighty-eight percent of respondents said it is becoming harder to tell whether online material is genuinely human or real. Eighty-four percent said convincing video no longer feels like proof, and 85% said scams are difficult to distinguish from legitimate offers; that figure rose from 66% the previous year.

Half of those surveyed reported experiencing some form of AI-related fraud or a scam, including being misled by AI-generated photos of products or receiving highly personalized scam messages. Nineteen percent reported specific identity-related harm linked to AI; about 10% said someone had used AI to create sexually explicit content of them without permission. Among people who use AI daily, roughly one in three said it is acceptable to generate explicit images of someone without their consent.

The survey found a gap between concern and protective action. Eighty-one percent fear someone could steal their family’s likeness, yet only 13% have created a family codeword to guard against impersonation. Sixty-seven percent worry about voice cloning, while 19% have disabled voicemail recordings to reduce risk. Respondents also reported comfort with some AI uses: 45% said AI assistance is acceptable for personal emotional writing such as wedding vows or eulogies, and 34% approved of using AI to create or improve a dating profile.

Survey participants gave concrete examples of harms and confusion. One parent described receiving a voicemail that sounded like their son asking for money for legal fees. Two unrelated respondents reported being fooled by the same AI-generated video of rabbits bouncing on a trampoline. Another respondent said an elderly relative showed them AI-generated material and believed it was authentic.

Malwarebytes released the findings in a report titled “Face value: How AI is reshaping trust, identity, and scams.” The report describes a rising presence of fake websites, products, images, voices and profiles and recommends steps users can take to reduce risk. It urges readers to “remember the human” and to consider tools such as scam detection and identity protection, plus simple practices like family codewords and limiting voicemail recordings, to guard against impersonation and fraud.