How the AI backlash will tank Western ambitions – and what can be done about it

The public backlash against AI is growing with support for AI collapsing at pace – rivalling even AI’s exponential capabilities curve. While the causes for this anti-AI sentiment are real and fully understandable, it creates in itself a major risk factor which could have catastrophic consequences.

In the UK, positive public opinion towards AI (defined here as “AI makes me excited” in Ipsos’ annual AI monitor reports), is now at just 29%, down from a high of 42% in 2024. The US and France both rank at 33%, both noting considerable drops from 2025.

This apathy is not a global phenomenon and is largely contained to the Anglosphere and Europe. Counterintuitively, it is the countries in the West with the most to gain from AI which have the least support for it.

It doesn’t have to be this way. In Japan, support sits around 47% with little year on year variance while China tops the list at 83%.

In this piece, we will explore why this backlash is happening, what the consequences are if it continues to go unaddressed, and how the fix is a fundamental shift in how we perceive AI.

Fuelling the fire

As a writer, LLMs threatened my craft before anyone else’s.

Realising this back in late 2022 was a long hard night of the soul and some harder whisky. Ultimately, game theory won out. The less time I procrastinated over pulling the lever, the more advantageous a position I’d be in.

Everyone has things beyond their control imposed upon them all the time – the difference between people is how they choose to respond. For me, I wasn’t exactly in the market for AI, but if you’re going to foist it upon me, you better believe I’m going to ride it like I stole it. I immediately went all in.

The reason I bring this up is to demonstrate that everyone responds to AI in a way that’s unique to them. The more naturally cautious will typically be more hesitant, the more carefree are pumping out workslop at a million tokens a second.

It is a mirror. How we perceive AI, our mindset going in, and our intentions all massively impact the outcome.

However, this mirror applies to more than just our individual reflections.

One area we see this acutely is Gen Z, who have massively turned against AI. On the surface, this sentiment is in stark contrast to what we’d typically expect. Younger generations have always been the drivers of the latest technology, and Gen Z were essentially born with a smartphone in hand.

Yet the data tells a very different story. A recent Gallup poll said that only 18% of Gen Z were “hopeful” about AI – dropping from 27% in 2025. Meanwhile, 44% are actively sabotaging AI initiatives in the workplace.

I’m only on LinkedIn and Reddit, but if either is reflective of the wider social media discussion on AI, then Gen Z certainly isn’t shy about vocalising this discontent either.

When you read what they have to say, the source of frustrations becomes clear.

Gen Z has essentially been the beta tester group of technologies which have left them more anxious, angry and isolated. They can’t buy a house without inheritance money, can’t start families, and any path to making it on their own saddles them with unbearable debts. The planet they stand to inherit is essentially a melted candle about to run out of wick.

It’s all the stuff millennials are pissed about, magnified.

Then along comes AI.

Degrees become essentially pointless. They can’t start a career as no one is doing entry level. Even when they do get a job, they learn nothing. The doors to anything creative are rapidly slamming shut. Already half lobotomised by social media, they are now expected to pick up AI with no proper training or encouragement, risking further braining themselves through cognitive offloading. One commentator said that “a world which already seemed fake and plastic takes another bold step away from reality.”

This is real pain, which neither the labs nor the nations hosting them have yet to offer a response to.

That pain is then compounded by another AI, the Payperclick Maximiser, turning social media outrage into trending content. The angrier you are about AI, the bigger the megaphone you get.

Muddying the waters further are China-backed AIs engaged in poisoning the well, pumping out anti-AI propaganda to increasingly receptive Western audiences. This isn’t something which is restricted to the CCP, however. Anyone with a mac-mini and an API budget (or, say, the biggest data centres on Earth, an axe to grind, and a trillion dollars in their backpocket) can influence our online hive mind.

It is a doom spiral, where anti-AI sentiment is promoted by social media’s AI algorithms, catalysed by anti-AI propaganda created by AI.

It is an irony, I am sure, that anti-AI people will say entirely proves their point – and they’d be right. If AI truly is a mirror, the reflection of what we see when we hold it up to society leaves a lot to be desired.

The strategy play

Given all the problems with the technology, many have called for its growth to be slowed – with even some calling for an outright ban.

This sentiment though causes many individuals and organisations to sit on the fence, or choose to not engage at all. Yet, as we’ve seen with the CCP bots, you are being engaged with. Non-participation is, in itself, is a major AI risk.

The crucial thing here is that cat’s already out of the bag. This is why we say AI’s more dangerous than nukes. There’s no snapping our fingers and making it go away. Even if the West were to outlaw the tech, it would continue regardless. The tech is out there and we must learn what to do with it.

Failure to do so won’t just see individuals or organisations getting left behind in a world. People who don’t adapt will drop out of contributing economically to society. Organisations that fail to react in time will get caught out by AI-native insurgents and go out of business. The resulting financial shockwave of both, I feel, is the real AI bubble.

Jeff Bezos sees this too, and his recently launched Prometheus is counting on it. The plan there is to buy up distressed companies with real world data which failed to make the AI jump, strip them for parts, and stick the data into a world model.

Crucially, the biggest risk of all is that the West fails to adopt AI. The best models ultimately mean nothing when everyone refuses to engage with them – particularly the generation expected to carry the tech forward. Far beyond just making Bezos et al richer when we fail, our ability to maintain technological and geopolitical relevance walks out the door when our people do.

Of course, this won’t convince someone against AI to perceive it in a different light. You could quite rightly look at the situation and conclude that a ‘leopard ate my face’ moment is much warranted.

Much of the problem comes down to how AI is communicated and framed.

We call it a tool, we positioned it as a jobs apocalypse, we’ve only ever really seen it as a product – something owned by billionaires, made to make endless more products with the goal of meaning we never have to even so much as blink.

Where is the vision in that? What is the strategic advantage? How does more data centres help people already in crisis and borrowing just to get by? Who is all of this for?

These questions continue to go unanswered in the West. Yet China’s excitement for AI shows there’s a different way.

The difference maker here – China’s use of AI to influence Westerners online notwithstanding – is that China positions AI is a strategy play, not product. There, the technology is positioned as core to the national interest and which will benefit its people. This fundamentally changes the game.

While the model wars rage and we obsess over API wrappers and shipping code, we’re failing massively at how to make it applicable and useful to everyone. This is also where the real advantage lies.

For example, the move to give the entire NHS access Copilot is a step in the right direction, but just handing it out is product-play mentality. The strategic play is enabling frontline healthcare workers so they understand how they can apply AI throughout their work. The Copilot version of AI in the NHS means less time on emails. The strategic vision is NHS workers with the ability to build AI workflows and tools bespoke to their individual needs to cut stress and burnout and improve patient outcomes.

The same is true of any sector – the unlock is empowering those closest to the biggest challenges.

Their success is entirely dependent on their relationship with AI. This means coming with the right mindset, approach, literacy and fluency, and intention. This is not something which can be achieved with product – they must want to do so. This is something which can only be delivered by strategy.

We must articulate clearly how AI benefits people, demonstrate how it can benefits the worker, and give them the agency to apply it. It is not something which can be achieved through force, but speaking to hearts and minds, and then delivering on the promise. If we fail to do this, we will fail to adopt AI.

Thanks for reading. I am currently writing a long form essay on why we need to see AI as strategy, not product, which expands on these themes and is due to be published shortly – hit subscribe if you want to see it.

I’m also about 2/3rds of the way through writing a book on AI strategy and adoption. At least I think it’s a book. It could also be a report, or something else entirely. I’m looking for people who can either contribute as sources or perhaps can help me steer it down a different medium (or just help publish it!). If you can help, please do get in touch.