Will The AI Bubble Burst in 2026?

Tech giants like Microsoft and Salesforce, along with management consulting firms like McKinsey and Deloitte, are laying off thousands of employees while claiming to reinvest in AI growth.

The news is prompting others to follow suit despite little evidence that automation is genuinely reshaping workforces at scale.

Late last year, the Revolut and Monzo-backed legal start-up, Robin AI, was suddenly listed for sale – shortly after appearing in the Sunday Times 100 Tech list. This begs the question: does AI really have the capabilities to take over human-led roles at scale?

The rise of the AI-driven layoff

All last year, tech companies pointed towards AI as their rationale for restructuring teams, with automation said be taking over certain jobs.

I don’t doubt that Amazon’s investment in AI-controlled robots and automation in its fulfilment centres is speeding up delivery times. But in the wider context, if AI is truly making roles redundant at pace, we’d see evidence across all sectors.

Instead, companies continue to hire junior employees, HR teams haven’t disappeared, and internal operations don’t appear to be lower staffed. The delta between what some companies are saying AI is doing, and what it’s actually doing, is clear.

The uncomfortable truth is many organisations have a vested interest in framing employment cuts as AI-driven, rather than as a response to market chill or reduced demand, to reinforce innovation credentials and build investor confidence.

When the world thinks AI is replacing thousands of jobs, they look futuristic and efficient. What we don’t hear about though, are the risks if strategic missteps are made.

The backlash could be brutal

Robin AI’s fall from rising star to potential acquisition could be a warning sign of things to come. Even in professional services, traditionally some of the most language-heavy, document-intensive industries, we are not yet seeing significant workforce disruption.

If fully AI-led services were ready to replace human professionals, continued investment surely wouldn’t be a challenge.

AI startups promise autonomy and scalability. Why bother engaging an expert when “there’s an App for that”? Why instruct an external adviser when one in-house expert can do the work of twelve?

I’m a fan of AI. The technology is remarkable and has incredible potential, but I don’t think it’s ready to replace professional advisers any time soon.

If large language models aren’t ready to replace people whose very stock-in-trade is language, then we’re far off replacing most other workers.

Jumping the gun represents a risk rather than a boon to productivity and service.

Risk and liability

Laying off staff on the assumption that AI can accurately and effectively replace them creates vulnerability if that’s not the case. And there isn’t only risk of losing those who are highly skilled in their professions for good. Graduates and apprentices will no doubt question whether prospective jobs will be safe in future.

Additionally, organisations that lean too heavily on automation risk exposing themselves to quality and liability issues. AI “hallucinations” (Latin for “making stuff up”) are well understood, and proponents argue that these problems are being dealt with.

But they’re only part of the challenge. Any experienced lawyer will tell you that reading a document and summarising what it says is the easy part. What it doesn’t say is often more important. And that’s before we get to nuance, sophistry and practical reality.

AI document review tools will apparently revolutionise the professional services market. I’ve seen these in action, and they’re impressive. Why pay dozens of juniors to review hundreds of documents if a tool can do it in seconds and flag potential issues?

The issue is appetite for risk. A client must be willing to make significant business decisions based on summaries of documents that nobody has ever read. This might be acceptable in certain scenarios, but ask yourself this: would you rather your surgeon actually reads your medical notes or takes a quick look at an AI summary before you go under the knife?

Where is the supposed productivity boom?

The promised efficiency gains simply aren’t materialising – at least not significantly enough to notice. AI companies repeatedly talk about extraordinary productivity leaps, but most businesses aren’t seeing the benefits yet. Partly because most jobs aren’t as simple as reading something and telling you what it says, and because mistakes and omissions matter.

The capabilities of AI shouldn’t be underestimated. There are (and will be) many useful tools. Software development and translation and transcription services are already showing the benefits of AI-assistance. The music industry is facing serious disruption. AI generated songs were a running joke this time last year. The Velvet Sundown and other AI music “projects” been the most sobering experience for music executives since cold water swimming.

However, I suspect that for most professionals, AI will be an innovation more like the calculator, the spreadsheet or computerised word processing: revolutionary, yes – enhancing human skills rather than replacing them. (The music industry will adapt too.)

As for the latest phenomenon, “Agentic AI” – where autonomous decisions are made to accomplish a set task with minimal human supervision – I’ll believe in the breakthrough when I see it in the wild.

I don’t doubt that improved AI tools, stronger quality checks and a clear-eyed understanding of the technology and its risks will enhance jobs over time.

What I do doubt is that AI will eliminate white collar professions any time soon. The rate of change is likely to be slower than AI boosters and CEOs would have you believe.

Will the bubble burst?

The familiarities with the early 2000s ‘.com’ bubble are becoming hard to ignore: overpriced valuations, investors backing the same horses, high expectations and start-up failures signalling fragility.

If more companies jump on the bandwagon – and AI can’t meet expectations – then 2026 could be the year the world becomes less starry eyed and a little more sceptical.

The technology will no doubt be transformative, but just not in the way many are being led to believe. Well, not yet anyway. Companies that continue to remain at the forefront will be those that use it strategically to enhance – not replace – human expertise and make their resourcing decisions based on what AI can do today, not what it might be able to do tomorrow.

Essentially, if AI isn’t going to be the gamechanger to the economy that it’s been hyped up to be and to generate the short to medium term returns that investors have been expecting, then… pop!

Published in Digit News on 23 January 2026.