The AI jobs impact may be far slower, broader and more complex than many headlines suggest. According to Microsoft vice-chair and president Brad Smith, artificial intelligence is unlikely to transform the labour market overnight, but its full effect could unfold over the next 25 years.
Speaking in Dublin at the International AI Summit, Smith argued that while mass displacement has not yet materialised, governments, businesses and workers should not become complacent. His message was clear: the long-term AI jobs impact will depend not only on technology itself, but on regulation, reskilling and how widely the benefits of AI are shared across society.
Why the AI jobs impact may take decades
Smith’s central point is that labour market change tends to happen in stages. New tools usually alter tasks first, then workflows, then entire job categories. That makes the AI jobs impact less like a sudden shock and more like a long transition.
So far, many companies are using generative AI to support workers rather than replace them outright. AI is helping with drafting, coding, research, summarising information and automating repetitive office tasks. In that sense, the current phase is more about augmentation than elimination.
However, that does not mean the risk disappears. Over time, roles may be redesigned, some positions may shrink, and new jobs may emerge in fields such as AI oversight, digital compliance, model training and human-machine workflow management.
Key reasons the transition may stretch over decades include:
- Slow adoption across regulated or traditional industries
- The need for businesses to redesign processes, not just buy software
- Skills gaps that limit how quickly workers can use AI effectively
- Government policy and labour protections that shape implementation
- Public trust and safety concerns around powerful AI systems
Skills and policy will shape the AI jobs impact
One of Smith’s strongest warnings was that societies need to invest now in helping people use AI productively. If workers receive the right training, the AI jobs impact could be more positive, boosting output and opening new career paths. Without that support, automation could widen inequality and leave many behind.
That puts reskilling at the centre of the debate. Employers may need to rethink workforce development in practical ways, including:
- Training staff to use generative AI tools safely and effectively
- Teaching critical thinking and verification skills
- Expanding digital literacy beyond technical teams
- Preparing managers to redesign jobs around human-AI collaboration
- Supporting mid-career workers facing role disruption
Smith also suggested that “creative” government policies will be needed. That could include tax incentives for training, updated education curricula, stronger social safety nets and targeted support for sectors most exposed to automation.
AI regulation is becoming a parallel priority
Alongside jobs, Smith made a forceful case for AI regulation. In his view, the capabilities of AI have advanced faster than the safety systems designed to govern them. That imbalance raises concerns not just for cybersecurity and misinformation, but for public confidence in AI adoption itself.
His argument is not for heavy-handed rules that choke innovation. Instead, he pointed toward practical, layered regulation that protects people while giving businesses clarity. Early rules may focus on incident reporting, accountability when AI systems cause harm and legal duties to notify victims in the event of cyber breaches involving AI agents.
This matters for the AI jobs impact because trust influences adoption. If businesses and consumers see AI as unsafe or unaccountable, deployment will slow. If regulation is balanced and predictable, companies may invest with more confidence and workers may be more willing to embrace the technology.
What sensible AI regulation could include
- Mandatory reporting when AI systems cause security incidents
- Clear liability rules for harmful autonomous actions
- Safety testing for high-risk AI applications
- Transparency standards around training and deployment
- International coordination on cross-border AI risks
Ireland’s strong position in the AI economy
Smith also highlighted Ireland as a country to watch in the AI transition. Based on Microsoft research, nearly half of the Irish population is already using generative AI, placing the country among global leaders in adoption.
That matters because countries with high usage rates may be better prepared for the coming AI jobs impact. A workforce that is already experimenting with AI tools is more likely to adapt as the technology becomes embedded in everyday work.
Ireland’s policy environment could also be an advantage. Smith praised the country’s long-term thinking and its ability to align national priorities with wider European regulation. For a small, open economy with a major multinational tech presence, that combination could make Ireland especially competitive in AI-enabled industries.
What workers and businesses should do now
If the AI jobs impact really will play out over 25 years, the takeaway is not to panic. It is to prepare. Businesses should focus on responsible adoption, workforce training and governance. Workers should build adaptable skills that complement automation rather than compete directly with it.
The most resilient qualities are likely to include judgment, creativity, communication, ethics and the ability to work effectively with AI tools. Technical skills will remain valuable, but human strengths will matter even more as automation expands.
In the end, the AI jobs impact will not be determined by algorithms alone. It will be shaped by decisions made today about education, regulation, workplace design and who gets access to the gains. The next 25 years may redefine work, but with the right planning, that change does not have to come at workers’ expense.
