A recent conversation with a client raised a question that I suspect many HR professionals will face in the coming years.
No employment tribunal has yet ruled on whether opposition to AI can be a protected philosophical belief. That is worth stating plainly, because the client conversation that prompted this piece made me realise how soon that question is likely to be tested.
The organisation was introducing a basic AI awareness session for employees. Nothing controversial. No requirement to use AI. No mandate to replace jobs with technology. Simply an introduction to what AI is, how it is developing, and what opportunities and risks it presents. The session would ask employees to draft a letter or email using AI and review it.
One employee objected.
Not because they were too busy. Not because they thought the training was irrelevant. But because they described themselves as philosophically opposed to AI, a self-styled “conscientious objector to AI,” and said it conflicted with their belief system.
My first reaction was that this was unusual. My second reaction was that perhaps it won’t be unusual for much longer.
Why is AI provoking this kind of resistance?
Throughout history, new technologies have sparked concern and resistance. The industrial revolution, mechanisation, computers, the internet and social media have all attracted critics who argued that the benefits came at a significant social cost.
Artificial intelligence appears to be generating a similar response, albeit on a larger scale. Concerns range from job displacement and surveillance through to environmental impact, misinformation, bias, and the longer-term implications of machines performing tasks previously undertaken by humans.
For some individuals, these concerns are not simply preferences or opinions. They form part of a deeply held ethical position about the role technology should play in society. That raises an interesting question: could opposition to AI eventually become a protected philosophical belief?
The legal test
Under section 10 of the Equality Act 2010, a philosophical belief is protected from discrimination if it satisfies the four-part test set out by the Employment Appeal Tribunal in Grainger plc v Nicholson (2009):
- The belief must be genuinely held, not adopted for convenience.
- It must relate to a weighty and substantial aspect of human life, not simply an opinion.
- It must attain a level of cogency, seriousness, cohesion and importance.
- It must be worthy of respect in a democratic society, compatible with human dignity, and not in conflict with the fundamental rights of others.
Protected so far: ethical veganism, humanism, pacifism, belief in man-made climate change, democratic socialism, and gender-critical beliefs (confirmed in Forstater v CGD Europe, 2021).
Not protected: support for a particular football team, Holocaust denial, claims that established terrorist attacks were staged, and ordinary vegetarianism, which was found to be a lifestyle choice rather than a weighty, coherent belief system.
No tribunal has yet tested opposition to AI against this framework. There was a time when ethical veganism and man-made climate change had not been tested either.
The possibility should not be dismissed out of hand.
Does attending AI training mean endorsing it?
No. There is an important distinction between requiring someone to understand a topic and requiring them to support it.
For employers, perhaps the more immediate question is not whether a belief is protected, but how they should respond when an employee raises such an objection.
In my view, most organisations already ask employees to sit through training on subjects they may not personally agree with or engage in: social media, cybersecurity, data protection, financial crime, political neutrality, or other emerging technologies. Attendance at training does not necessarily amount to endorsement, and UK employees are generally expected to follow reasonable and lawful management instructions, which typically includes role-relevant training. Refusing without a legally protected reason can be treated as a failure to follow a reasonable instruction, though any employer should still investigate the reason for a refusal, consider reasonable adjustments, and follow the ACAS Code of Practice before treating it as misconduct.
Understanding the impact of AI on the workplace is rapidly becoming a business necessity, regardless of whether an individual intends to use it.
Is this a new kind of diversity challenge?
Perhaps the most interesting aspect of this debate is what it tells us about the future workplace.
For years, organisations have focused on diversity of background, experience and thought. AI introduces another dimension: diversity of technological philosophy.
Some employees are enthusiastic adopters. Others are cautious. Some are openly sceptical. A small number may reject AI altogether on ethical grounds.
As employers, the challenge is unlikely to be choosing one side or the other. Instead, it may be creating workplaces where employees can hold different views while still engaging constructively with the reality that AI is becoming part of modern working life. That challenge is only becoming more pressing: a government-backed survey published in 2026 found that roughly one in three UK adults had used AI at work in the past month, while the large majority of the workforce had received no formal AI-related training in the previous year.
Where does this leave employers?
It leaves them managing something they haven’t had to manage before: not just the technology itself, but employees’ beliefs about it.
The employee who objected to the training may ultimately prove to be an outlier. Or they may represent the beginning of a much wider debate.
As AI becomes increasingly embedded in recruitment, performance management, customer service, learning and development, and everyday decision-making, organisations will need to consider not only how they implement AI, but how they respond to employees who have principled objections to it. That is a challenge HR has not faced before.
FAQs
Q: Is opposition to AI a protected philosophical belief in the UK?
A: Not yet. No employment tribunal has ruled on the question. To succeed, it would need to satisfy the four-part test from Grainger plc v Nicholson (2009): genuinely held, relating to a substantial aspect of life, cogent and serious, and worthy of respect in a democratic society.
Q: Can an employee legally refuse mandatory AI training?
A: Generally, no, not without a legally protected reason. Refusing a reasonable, lawful management instruction such as role-relevant training can be treated as misconduct, though employers should investigate the reason for refusal and consider reasonable adjustments before taking disciplinary action.
Q: Does requiring an employee to attend AI training mean the employer endorses AI?
A: No. Requiring understanding of a topic is different from requiring support for it. Employers already run mandatory training on subjects employees may not personally agree with, such as data protection or political neutrality, without that amounting to endorsement.
