Expert: Telling real content from AI-generated is becoming nearly impossible

Guna Puce, head of the Latvian Artificial Intelligence Center, said on the TV24 program "Uz līnijas" that it is no longer possible to reliably tell whether information found online is real or created by artificial intelligence. According to her, this applies not only to video and images but especially to text content.

Technology outpacing detection

Puce noted that until recently, deepfakes could be identified by certain telltale signs, but rapid technological advances have largely eliminated that possibility. She believes people should expect that in the future it will no longer be possible to confidently distinguish genuine content from AI-generated material.

Consumption habits need to change

Since content can no longer be recognized "by eye," the expert stresses that attention must shift to how people consume information in general. She warns that emotionally charged content designed to trigger an immediate urge to share is often created precisely for that purpose — so people spread it without checking it first.

Recommendations for users

Puce advises thinking about who published a piece of information and why one feels the urge to share it before doing so. For shocking events, confirmation should be sought from official news and media sources rather than relying solely on what is seen on social networks.