Is the internet starting to feel less human?
Polling shows a majority of the public now struggle to tell what’s real online, with growing consequences for trust and behaviour.
Do you ever get the sense the internet feels less real than it used to? Increasingly, the people we interact with online are not always people at all. Bots, automated accounts and synthetic content now make up a growing share of what we see, read and respond to. New polling suggests the public is clocking this shift, and that it is already changing how people experience the internet and how much they trust it.
New polling from Public First has found:
Over half of UK adults (55%) say they have encountered bots, synthetic media or deepfakes in the last year.
More than two thirds (68%) say they have mistaken AI-generated content for something real or human - over Christmas, 35% of the public say they encountered AI-generated scams.
Over seven in ten (71%) say they are now more cautious about clicking links or buying things online.
Synthetic media, bots and deepfakes are becoming, part of everyday internet use
Bots, synthetic media and deepfakes, often generated using AI are now a routine feature of the British Public’s daily online experience.
Over half of UK adults (55%) say they have encountered bots, synthetic media or deepfakes in the last year. Around two thirds (65%) of those who have encountered AI online say they have come across bots, compared with 42% for synthetic media and 41% for deepfakes.
However, this experience is most common among younger users. 71% of 18–24 year olds say they have encountered bots, synthetic media or deepfakes in the last year, compared to just a third of those aged 65 and over. This gap likely reflects differences in platform use, but may also reflect differences in confidence identifying AI-generated material.
Many people struggle to tell what is real
As AI tools have become more sophisticated, distinguishing between human and AI-generated content has become harder. More than two thirds (68%) say they have encountered AI content they initially believed was real or human, but later discovered was not.
For most people, this is not a rare occurrence. Nearly three quarters (73%) of those who have been mistaken say this happened within the last month.
There are clear age differences here too. 61% of people aged 65 and over say they have never been fooled by AI-generated content, compared with just 11% of 18–24 year olds. This may reflect differences in exposure, but it may also point to differences in awareness or confidence when navigating online environments.
Scammers have added synthetic media to their arsenal
Digital scams are on the rise in Britain. Nearly two thirds (64%) believe digital scams and fraud have increased over the past year. This perception is reflected in official data: the National Crime Agency identifies fraud as the most prevalent crime against individuals in England and Wales, with 67% of reported cases now cyber-enabled.
For many of the public, this is linked to an increasingly non-human internet. More than half of UK adults (55%) say they have received a message or call they believe was AI-generated to trick them. One in three encountered an AI scam in December alone, and 6% say they fell victim to one. Overall we estimate 20 million people could have been exposed to an AI scam over Christmas.
These experiences are already translating into financial harm. More than a quarter of UK adults (27%) say they or someone they know lost money due to an online scam or misleading content that may have used AI. Among 25–34 year olds, this rises to nearly half (47%).
Looking ahead, the public expects these challenges to intensify. Around two thirds (68%) think AI will increase the volume of scams and fraud, while just 4% believe it will make it easier for law enforcement to stop them.
Brits are changing how they behave online
As a result of the increasing scams and difficulty identifying real humans as opposed to bots or deepfakes, many users are adapting their behaviour.
As trust erodes, people may retreat from parts of the internet that once enabled easy connection and commerce. Four in ten UK adults (40%) say they are using certain online services less because bots and AI-generated content have damaged the experience.
Over seven in ten (71%) say they are now more cautious about clicking links or buying things online. This has major implications for economic activity online, particularly for smaller firms and creators that depend on low-friction digital trust, without being brands that are household names.
How bots, synthetic media and deepfakes could challenge the fundamentals of the internet
Bots and automated content are now embedded across much of the internet. In many cases, they serve useful purposes. But our polling suggests their growing presence is also changing how people experience being online.
Without clearer signals of authenticity, of who is real and who is not, alongside stronger protections against non-human misuse, trust in the internet is likely to continue to erode. The risk is not just more scams or worse content, but online spaces that feel less human and less worth engaging with, encouraging people to disengage from shared online communities. At the same time, reduced trust threatens economic activity online, particularly services and businesses that rely on confidence between strangers.
Major technology companies are increasingly engaging with these challenges. Google, Microsoft and Meta are deploying AI-driven systems to detect scams, impersonation and bot activity, and are working through initiatives such as the Coalition for Content Provenance and Authenticity to develop shared standards that help users understand where content comes from and whether it is real. Alongside this, Tools for Humanity is developing privacy-preserving digital identity tools to help prove a user is human online, reflecting a growing recognition across industry that rebuilding trust requires coordinated action.
Written by Olly Stanton