Share
Artificial intelligence and automated traffic may have officially overtaken human activity on the internet.
But a new cybersecurity study has exposed a far more alarming reality: almost half of us can no longer tell the difference between a real person and an automated bot online.
A week-long public experiment titled “Bot or Not?”, conducted by cybersecurity firm Surfshark in collaboration with master’s students from Malmö University, revealed that 47% of people fail to spot bots on social media accurately.
The study analyzed data from 710 participants during Milan Design Week, tasking them with separating genuine human accounts from automated ones.
The findings indicate that digital radar is heavily influenced by demographics, platform preferences, and, most notably, emotional engagement.
“We can’t simply ‘read’ our way out of ‘botted’ social media,” warned Luís Costa, Research Lead at Surfshark. “The most striking finding was that our biggest blind spot is emotion: when debates get heated, it hijacks our digital radar.”
According to the data, age remains one of the strongest indicators of digital literacy. Young people under the age of 20 proved to be the most resilient, boasting a 71% accuracy rate when identifying automated accounts.
While performance remained steady throughout participants’ 20s and 30s, bot-detection skills plummeted for individuals over the age of 40. Furthermore, the study noted that users who primarily browse Facebook or Threads, alongside people who avoid social media entirely, are statistically the most likely to fail the test.
Crucially, the experiment exposed how sophisticated bot farms manipulate human psychology. When analysing dry, non-emotional technical subjects, such as data centres, participants successfully spotted 71% of the bots. However, when conversations turned toward highly polarised political or social issues, human detection skills degraded rapidly.
On topics such as immigration, bot detection dropped to 54%, with users becoming increasingly paranoid and frequently misidentifying real human accounts as bots. When testing sensitive topics such as women’s rights, human accuracy hit its lowest point at 49%, meaning participants missed more automated accounts than they caught.
Cybersecurity experts warn that generative AI feeds directly on impulsive responses. To stay safe, internet users are urged to slow down before reacting to viral content, treat unsolicited direct messages with caution, and remain highly sceptical of emotionally manipulative posts designed to spark immediate outrage.
For complete research material behind this study, click here.
Related Posts
Discover more from Tech Digest
Subscribe to get the latest posts sent to your email.
Chris Price
Source link