Artificial intelligence promised to transform business. In 2026, the bill is arriving. Across Ireland, the UK, South Korea, and the EU, a more sober picture is emerging: high costs, mixed returns, reshuffled job markets, and a new regulatory layer that changes how companies can deploy AI at all.
Irish Businesses Are Losing Money on AI
A Censuswide survey of senior IT executives in Ireland, conducted in March 2026 and cited by The Irish Independent, found that Irish organisations have collectively lost €720 million on AI investments where returns failed to materialise. The phrase now gaining traction in tech circles is “token thrifting” – a sharp reversal from the earlier “token maxxing” era, when companies competed to pour resources into AI tools regardless of measurable output.
The pattern is not unique to Ireland. Across ordinary businesses, AI costs are rising faster than the value being generated. That raises real questions about whether the productivity gains many leaders anticipated are actually being delivered.
South Korea’s AI Stock Bubble Has Burst
The Financial Times recently published a headline that captured the mood among retail investors who piled into AI-linked stocks: “‘My life’s screwed’: Korean investors stress out after AI bubble bursts.” As reported by The Irish Times on 2 August 2026, South Korea’s experience is being read as a lesson the wider market has seen before. Speculative enthusiasm outpaces fundamentals, and retail investors typically absorb the losses when sentiment reverses.
It is, as the article notes, an old market lesson. AI is not exempt from hype cycles.
The EU AI Act Is Now in Force
August 2026 marks a significant regulatory moment. New rules under the EU AI Act – passed by the European Union in 2024 and described as the world’s first comprehensive AI law – have now come into effect. Among the most immediate requirements: companies operating in the EU must disclose when a user is interacting with an AI chatbot rather than a human.
In Ireland, the government has established the AI Office of Ireland to oversee compliance. Niamh Smyth, Minister for Artificial Intelligence, described it as “an important milestone.” Separately, a UK petition is calling for greater transparency from AI companies about what data was used to train their models – a sign that public pressure on AI accountability is building beyond the EU’s borders.
Jobs: Senior Roles Are Growing, Entry-Level Roles Are Not
According to reporting by Yahoo Finance, AI is creating a two-speed jobs market in the UK. Demand is rising for senior technical roles – people who can build, manage, and govern AI systems. At the same time, entry-level positions in tech and finance are being cut. The same pattern is visible in Ireland, where junior workers appear to be among those most exposed to AI-driven displacement.
This is not a forecast about the future. It appears to be happening now, in 2026.
Key Takeaways
- High AI spending is not automatically producing returns. Irish organisations have reportedly lost €720 million where ROI has not matched costs, according to a March 2026 survey of senior IT executives.
- Retail investors in AI stocks face real risk. South Korea’s experience illustrates that speculative excitement around AI is not a reliable investment signal.
- The EU AI Act is live. Companies must now disclose AI chatbot interactions and meet new compliance obligations. Ireland has established a dedicated AI Office to oversee enforcement.
- The labour market impact is uneven. Senior tech roles are in demand while entry-level positions in tech and finance face significant pressure in both the UK and Ireland.
- Transparency demands are growing. Calls for disclosure on AI training data are building in the UK, suggesting the regulatory environment may extend further in the near term.
Where This Leaves Us
None of this makes AI irrelevant. It does make it more accountable. The businesses and investors most likely to navigate this period well are those treating AI as a tool that requires careful cost-benefit analysis, not a guaranteed competitive advantage. The hype cycle has not ended AI’s potential – it has just started asking AI to prove it.
Photo by Helena Lopes on Unsplash