From bubbles to 'SlopTok': Why 2026 could see both the highs and lows of artificial intelligence

I look at the prospects for artificial intelligence (AI) in 2026 with a cloud of unease smattered with faint glimmers of hope.

We are in an AI bubble, with Morgan Stanley estimating almost $3 trillion will be spent buying computer chips and building data centres by 2028, but relatively little revenue arising from AI-based services to pay back the debt financing those investments.

The Bank of England recently warned that a “sharp correction” could mean a market crash similar to the 2008 financial crisis. However, the AI industry may be considered ‘too big to fail’: governments around the world, including here in the UK, are betting big on AI-driven productivity and economic growth. It seems more likely they will seek to reassure investors by offering incentives and long-term contracts; reducing barriers to data centre construction; and retaining a low-regulation environment.

Read more:

• Big tech wants to mine human creativity for AI training. We say no
• Labour wants to use ‘Minority Report’ AI to predict murder. This could have chilling consequences
• How do we stop AI becoming an ‘engine of inequality’, and make it a force for good?

If the bubble doesn’t burst, we will still be trapped in it, along with AI firms increasingly desperate to bring in revenue (and stay on the right side of Trump) regardless of the impact on our information environment. We are already subject to low quality AI-generated websites and dodgy search summaries. OpenAI’s Sora 2, which some have dubbed “SlopTok” has been used to churn out deep fake videos that stoke racial tension. Elon Musk’s Grok has been used to create a version of Wikipedia riddled with conspiracy theories. AI is accelerating social media’s assault on our shared understanding of the world.

We will also see increasing evidence of the negative impacts of AI on people’s lives. I expect further examples will surface of emotional and cognitive reliance on AI chatbots, particularly amongst young people. No doubt there will be more reports of what some call AI psychosis, where chatbots engage in a folie à deux with their users, leading in extreme cases to mental breakdowns and suicide. AI business models incentivise dependence-by-design.

AI’s impact on work will also increase. The creative sector may be most obviously hit by job losses, but there may also be fewer opportunities for graduates in professions such as law, consulting and software development. And it’s not just job quantity that could suffer: increased automation may also increase workplace surveillance and lead to work intensification, and reduced job satisfaction, in ways that are harder to measure than unemployment rates.

We will also see further out-and-out errors happening in places where AI, and particularly AI agents, are used without proper testing against the messiness of the real world. Reports will continue to surface of data leaks and financial losses; cyber attacks and deep fake scams; energy demands and water shortages.

If all of this sounds grim, it’s because I expect 2026 to be a nadir: the rock bottom we need to hit in order to get a better settlement around AI. Seeing and experiencing the worst destructive excesses of the AI industry may be what it takes to generate the popular will and political determination to demand something better.

I expect more of us to be, like Dame Emma Thompson, infuriated by AI being mainlined into our veins, as prime minister Starmer put it. I think we need to see the realities of AI slop to help us appreciate human creativity, just as experiencing the limits of artificial workers will help executives (and customers) appreciate their human workforce. The multiple lawsuits OpenAI now faces for contributing to people’s delusions and suicides may prompt calls for regulations and safeguards that prevent general-purpose chatbots being used as coaches or counsellors. The government’s recent mistakes detecting child benefit fraud may inject necessary caution into the roll-out of future automated systems.

But it will take action in 2026 and beyond to set us on a better path. It will take unions and their members negotiating the use of AI in the workplace. It will take parents, teachers and support staff speaking up to shape its adoption in schools. It will take customer complaints and boycotts to influence the market; letters, petitions and rallies to prompt government action. It will take each of us discussing the role of AI in our lives with our families, friends and colleagues to determine the future we want to work towards.

So while the outlook for AI in 2026 looks bleak, it could also bring something we need: the urgency to act. If we can channel our frustrations with AI into organising, regulating, and building something better, we could still bend the trajectory of this technology toward something genuinely beneficial.

Jeni Tennison is the executive director of Connected by Data.

Do you have a story to tell or opinions to share about this? Get in touch and tell us more. 

Reader-funded since 1991 – Big Issue brings you trustworthy journalism that drives real change.

Every day, our journalists dig deeper, speaking up for those society overlooks.

Could you help us keep doing this vital work? Support our journalism from £5 a month.