Artificial intelligence has acquired an unexpected social stigma: admitting to using it can now be almost as damaging as producing something badly written. Across politics, workplaces and public life, the use of AI is increasingly treated not simply as a question of competence or judgement, but as evidence of intellectual laziness, professional inadequacy or even moral failure.
A recent Mail on Sunday headline, declaring that “ROBOTS REALLY ARE TAKING OVER!”, illustrated the mood. Its target was an aide to Andy Burnham who appeared to be consulting ChatGPT while reviewing a draft of a speech for the prime minister on a train to Holyhead. The apparent offence was not that the aide had failed to do his job, but that he had used an artificial intelligence tool to help perform it. The episode became another example of what is emerging as a culture of public hostility towards those who acknowledge using AI.
The phenomenon is not confined to Britain. In April, a provincial minister in Newfoundland, Canada, broke down in tears in the House of Assembly after facing days of criticism over her use of AI to alter a photograph of a provincial art gallery for a social media post. In Sweden, Prime Minister Ulf Kristersson was ridiculed and criticised in August 2025 after revealing that he regularly used ChatGPT and LeChat to assist his official decision-making. He described the tools as a means of obtaining a second opinion, testing alternative approaches and considering what others might have done.
In the United States, the social penalties appear equally pronounced. A Duke University study found that participants consistently regarded employees who used AI as lazier, less competent, less diligent and less independent than colleagues who relied on other people for assistance. The findings suggest that the stigma attached to AI use is no longer merely an expression of technological scepticism. It is becoming a judgement about the character and abilities of the people who use it.
The American legal scholar Lawrence Lessig has identified a growing tendency towards moral condemnation of AI users. As more people incorporate the technology into their professional and personal lives, he argues, sections of society have developed a self-righteous impulse to demand that users disclose their reliance on it. The result is an emerging expectation that using AI must be publicly acknowledged, almost as though it were an embarrassing professional secret.
This suspicion has helped create a market for AI-detection services, which claim to identify writing produced by artificial intelligence. Yet their reliability remains questionable. One particularly revealing case involved an Open University academic whose entirely human-written article was rejected by a specialist journal after an AI detector classified it as machine-generated. The system reportedly cited the article’s unusual fluency, consistency, uniformity and lack of syntactic irregularities as evidence against its authenticity. In effect, the very qualities expected of competent writing became grounds for suspicion.
Such cases expose a fundamental problem with attempts to police AI use through detection. If coherent prose, consistent sentence structures and grammatical accuracy can be interpreted as evidence of artificial production, the distinction between human and machine writing becomes dangerously unreliable. The technology intended to expose deception can itself generate false accusations.
But the more troubling dimension of AI shaming concerns social inequality. As the argument presented in Observer UK suggests, hostility towards AI use can amount to a form of professional snobbery, reinforced by a striking double standard. Political leaders, senior executives and public figures have long relied on assistants, speechwriters and communications teams to prepare their speeches, statements and articles. Their dependence on other people’s intellectual and writing skills is generally accepted as an ordinary feature of professional life.
Yet an employee who uses ChatGPT to improve a document may be regarded as intellectually dishonest or incapable of doing the work independently. The difference is not necessarily the quality of the final product, but the social status and resources of the person producing it. Those with money and influence can obtain human assistance without attracting scrutiny. Those without access to such support may turn to AI and find themselves publicly judged for doing so.
The consequences are likely to be uneven. People with limited educational opportunities, those who write in a language that is not their first, and neurodivergent individuals who experience difficulties with writing may have particular reasons to use AI as a practical aid. Treating every instance of AI assistance as evidence of incompetence risks penalising precisely those who could benefit most from access to the technology.
That does not require accepting every product generated by artificial intelligence. Poorly checked, generic and inaccurate AI output deserves criticism. Nor should the technology replace tasks whose educational purpose is to develop independent thought and writing skills. Students, for example, need opportunities to formulate arguments, test ideas and discover their own intellectual positions through the process of writing.
But not every writing task serves that purpose. Much professional and administrative communication consists of routine, bureaucratic prose, where the principal requirement is clarity rather than the development of an original intellectual position. In such circumstances, using AI to produce or improve a document may be no more objectionable than relying on a secretary, editor or communications assistant.
The central distinction, therefore, should not be between those who use AI and those who do not. It should be between responsible and irresponsible use. Accuracy, transparency where necessary, accountability and the ability to assess the quality of the final work matter considerably more than whether a person has employed an artificial intelligence tool along the way.
Public shaming is unlikely to reverse the spread of AI. It may instead encourage people to conceal their use of the technology, making meaningful disclosure, scrutiny and the development of professional standards more difficult. A culture in which employees fear admitting to AI assistance is hardly conducive to establishing responsible practices.
The deeper question is why a technology that can make writing and administrative work more accessible has become a test of professional respectability. If influential people can employ teams of assistants without being accused of intellectual laziness, it is difficult to justify condemning ordinary workers for using a tool that performs some of the same functions.
The most revealing response to someone who asks whether you use AI may therefore be the simplest one: of course. The more pertinent question is why they believe they should not.

