Over the past two years, artificial intelligence has moved from academic research labs and industry testbeds into the everyday lives of millions of people. Nowhere is this transformation more visible—or more consequential—than in the rise of conversational AI systems such as ChatGPT. These tools generate text, answer questions, draft emails, summarise articles, and participate in online conversations at a scale that was difficult to imagine even a decade ago. For many users, they have become an integral companion at work, at home, and across their digital interactions.
Yet it is on social media and within the broader digital public sphere that ChatGPT’s influence is most profound. Here, ideas are exchanged, opinions are shaped, and narratives spread with unprecedented speed. The introduction of a system capable of generating persuasive, human-like language raises pressing questions: How is ChatGPT affecting online public opinion in the UK? Are we witnessing the beginning of a new era in democratic discourse, or merely the acceleration of trends already underway? And crucially, how should we—as a society—respond?
This commentary aims to address these questions from the perspective of a member of the UK academic community, speaking not only to experts but to the general British public. My aim is to offer clarity, nuance, and a grounded assessment of ChatGPT’s influence on social media and online opinion formation, at a time when such analysis is urgently needed.
To appreciate the magnitude of ChatGPT’s impact, we must understand the broader context of our digital information environment. Social media platforms such as Facebook, X (formerly Twitter), TikTok, Instagram, Reddit, and YouTube already served as the primary ecosystem through which many Britons—especially younger demographics—consume news, express views, and engage in public debate.
Unlike previous technological advances, ChatGPT did not gradually infiltrate this environment. It arrived fully formed, free to access, and instantly capable of generating vast volumes of coherent text indistinguishable from human writing. The result was a sudden shift in the way information could be created, shared, and amplified.
Within weeks of ChatGPT’s release, social media analysis firms reported dramatic increases in AI-generated posts. These ranged from harmless productivity tips and motivational messages to more problematic content:
automated political commentary
fabricated expert advice
AI-generated news snippets
influencer scripts
automated replies to trending hashtags
AI-driven “debates” between user bots
The ease of generation fundamentally altered the economics of online content. What previously required time, skill, or financial investment could now be produced instantly.
The internet once amplified human voices; today it amplifies synthetic voices as well. On platforms where quantity often trumps quality, ChatGPT became a force multiplier. For individuals, it simplified content creation. For organisations or coordinated groups, it enabled large-scale messaging campaigns at minimal cost.
This change has made the digital public sphere simultaneously more vibrant and more volatile.
ChatGPT’s presence in online discussions is not inherently harmful. In many ways, it has introduced benefits that support public learning, increase access to information, and democratise knowledge. Yet the very capabilities that enable these benefits also introduce risks.
Many Britons—especially those less familiar with technical jargon, complex policy debates, or specialised academic language—have found ChatGPT to be a valuable mediator. It can:
explain political concepts in plain English
summarise parliamentary debates
clarify scientific findings
provide context for breaking news
translate policy documents into accessible language
This represents a democratising force in public discourse, empowering individuals to engage more meaningfully with topics previously perceived as opaque or inaccessible. Teachers, journalists, parents, and professionals often use ChatGPT as an interpretive tool, bridging the gap between expert knowledge and the broader public.
However, ChatGPT is not a journalist, a scholar, or a policymaker. It does not possess inherent truth; it generates plausible language based on patterns. This means:
It can be confidently wrong.
It can inadvertently spread inaccuracies.
It can produce fabricated sources, quotes, or statistics.
It may fail to recognise context, irony, or emotional nuance.
On social media—where users often skim content rather than scrutinise it—these limitations can lead to the circulation of misinformation.
Public opinion is not a static entity; it evolves through exposure, conversation, persuasion, and social norms. ChatGPT affects each of these processes in distinct ways.
One of ChatGPT’s most significant influences is its capacity for personalisation. Instead of one public message broadcast to millions, individuals now receive individually tailored responses shaped by their prompts, concerns, and prior beliefs.
This is both empowering and potentially dangerous. Personalisation can:
help clarify complex issues
create a sense of trust and rapport
lead users to feel heard and understood
But it can also:
reinforce confirmation bias
amplify partial or one-sided interpretations
create fragmented pockets of understanding
In other words, ChatGPT may inadvertently contribute to a more divided digital public sphere.
Many users assign undue authority to ChatGPT due to its articulate tone, coherence, and the prestige associated with advanced technology. The danger lies not only in misinformation but in the shifting dynamics of trust.
A generation raised on search engines often trusted information indexed by algorithms. Today, a significant portion may place trust in language generated by algorithms. When authority becomes algorithmic, responsibility becomes diffuse.
Political actors have already begun using AI for:
drafting speeches
generating campaign messaging
simulating audience reactions
creating tailored ads
producing social-media-ready content at scale
In the UK, this raises concerns about:
the integrity of online campaigning
transparency in political messaging
the risk of automated “astroturfing”
synthetic public support swaying undecided voters
Although regulatory safeguards such as the Electoral Commission’s guidelines on digital campaigning have improved, AI’s speed and adaptability challenge the pace of policy.
Social media thrives on virality—speed, scale, and emotional resonance. ChatGPT affects each of these drivers.
Posts that previously took hours to craft can now be produced in seconds. This amplifies:
advocacy campaigns
ideological narratives
trending topics
disinformation cascades
The faster content spreads, the less time users have to verify accuracy or reflect critically.
Automated posting, even when not malicious, can overwhelm organic discourse. This skews perception of public sentiment by creating an illusion of consensus—or conflict—where none exists.
ChatGPT can systematically generate emotionally engaging content. Emotion drives engagement; engagement drives visibility. This increases the likelihood that AI-generated posts will trend, reach new audiences, and influence debates.
Trust—once eroded—cannot be easily restored. ChatGPT’s role in online discourse raises critical questions:
How can users discern human opinions from AI-generated ones?
Should platforms label AI-generated content?
How might public trust evolve when digital voices become increasingly synthetic?
While organisations like OpenAI commit to safety and transparency, responsibility also lies with:
government regulators
academic institutions
media organisations
digital literacy advocates
individual users
No single entity can safeguard the digital public sphere alone.
If users perceive online discourse as artificial or manipulated, they may withdraw from civic participation. This “civic fatigue” can weaken democratic debate and undermine public confidence in institutions.
The UK is uniquely positioned to shape the future of AI governance. Our academic community, regulatory institutions, tech sector, and public broadcasters all play pivotal roles.
Just as schools introduced media literacy to navigate the early internet, Britain now needs AI literacy to understand:
how AI generates content
its limitations
how to verify sources
how to recognise AI-mediated persuasion
This is not merely educational; it is democratic.
Policy interventions may include:
mandatory labelling of AI-generated political content
auditing automated messaging campaigns
regulating synthetic media in elections
transparency requirements for AI-assisted journalism
These require careful balancing to protect free expression while safeguarding the integrity of the digital commons.
ChatGPT should not be framed solely as a threat or as a panacea. It is a tool—one that reflects and amplifies human intentions. The task ahead is to harness its benefits while mitigating its risks.
ChatGPT has already reshaped online discourse in ways profound and irreversible. It empowers individuals, enhances access to knowledge, and enriches public debate. Yet it also introduces ambiguities, vulnerabilities, and new ethical dilemmas.
Britain now stands at a crossroads. We must decide whether to allow the digital public sphere to evolve without guidance—or to shape it intentionally, thoughtfully, and democratically. The key lies not in resisting AI but in adapting our norms, expectations, and institutions to the realities of algorithmic conversation.
The future of public opinion will not be written solely by humans or machines, but by the interplay of both. Our task is to ensure that this interplay strengthens, rather than undermines, the foundations of a healthy democratic society.
