AI, Growth, and Jobs, Who Really Benefits?

People love a neat AI story. Either AI will take everyone’s job, or AI will make us all so productive we can retire early and become sourdough influencers. Reality is messier, and much more useful for leaders.

A recent piece republished by ScienceAlert makes a strong point: the biggest risk is not mass unemployment, it is how the gains from AI get shared. That plays out in a few clear ways inside real organisations. If AI makes some people more valuable while making others more disposable, the business question becomes less about panic and more about choices.

Here, we break down the jobs-and-growth story into three practical angles, then adds the missing piece: distribution and skills determine who wins.

• The scale effect
• The new products effect
• The SME effect
• The distribution effect you cannot ignore

1. The scale effect

Productivity creates capacity, capacity creates options

If AI reduces the time and cost of delivering work, you free up capacity. What happens next is not automatic. It depends on what you do with that capacity.

There is evidence from a Bank of England report that firms expect (and are seeing) productivity gains, but they also expect labour impacts to be gradual because systems need integrating and outputs often need human checking. That lines up with what many leaders experience: the first wave feels like faster work, not fewer people.

The growth step happens when productivity gains get reinvested into expansion. Research by Tania Babina and co-authors finds AI-investing firms have higher growth in sales and employment, but job creation is not automatic. It depends on what the firm does next with the extra capacity.

So two firms can adopt the same tools and get opposite outcomes.

• Firm A takes the savings, freezes hiring, and banks the margin.
• Firm B reinvests the saved time, takes on more work, and hires in delivery, operations, customer success, and sales.

A practical question that forces clarity is this: what will you do with the capacity you free up?

• If the answer is nothing, you have chosen a cost-cutting path.
• If the answer is growth, you have chosen a more jobs-friendly path.

2. The new products effect

The biggest impact is not faster admin, it is new offers

AI’s job impact often shows up downstream.

When drafting, analysing, support triage, and routine queries get cheaper, firms can change what they sell, not just how fast they work. That is where growth, and therefore hiring, tends to come from.

Babiba et al point to product innovation as a primary driver of growth for AI-investing firms. UK evidence, again from the Bank of England, also highlights a wide range of business processes where AI is being applied, including customer service support and automation of routine tasks, which can enable new or improved services.

This is why the hiring pattern usually shifts away from AI job titles and towards roles that make the new offer real, safe, and repeatable.

• Implementation and onboarding
• Quality assurance
• Compliance and governance
• Data governance and information management
• Advisory-led sales and account management

This is also where the ScienceAlert framing helps. It argues technology can raise job value by shifting humans into work customers care about, using the ATM example where teller roles moved towards service and advice rather than pure cash handling.

A better question than “Will AI replace me?” is this : Will AI make me more productive in a way customers value? If yes, you become more valuable. If no, you risk becoming a cost line item.

3. The SME effect

The real limit is time and skills, not ambition

Big-business AI debates get the headlines, but SMEs live in a different world. The constraint is rarely ideas. It is time, skills, and headspace.

OECD research finds generative AI is already being used in a meaningful share of SMEs. SMEs report improved performance and help with skill gaps and labour shortages, while also reporting increased demand for highly skilled workers.

That combination is the reality check.

• AI can remove bottlenecks and lift output.
• Growth still needs humans, especially for judgement, accountability, and trust.

What this looks like in practice.

• A small accountancy firm uses AI to cut admin and launches a monthly advisory tier, then hires a client manager earlier than planned.
• A trades business uses AI for quoting and aftercare, wins more work, and takes on an apprentice.
• A training provider uses AI for personalised practice and course content, improves outcomes, and needs more trainers and facilitators, plus extra capacity for booking, learner support, and admin.

AI does not create jobs in the abstract. It removes the capacity ceiling that was stopping the business from taking on more work.

4. The distribution effect you cannot ignore

Unequal gains are the risk, and leaders influence this

A big risk is that some people benefit far more than others, partly because using advice and tools well is itself a skill.

The IMF makes this point: AI can lift productivity, but it can also widen inequality if the benefits accrue mainly to those who can harness the tools.

There is a second uncomfortable truth: some firms may use “AI transformation” as cover for standard cost-cutting.

Real-world examples of how this shows up.

The pattern is cutting headcount before redesigning the work.

• A marketing team adopts AI content tools, then cuts junior roles instead of using the extra capacity to test more campaigns and improve performance.
• A customer support team rolls out an AI chatbot, then reduces human cover hours while keeping the same response-time targets.
• A back-office team automates reporting and admin, then removes roles without redesigning processes, leaving the remaining staff to manage exceptions and errors under more pressure.
• A professional services firm uses AI for first drafts, then trims headcount while still billing for “efficiency”, rather than reinvesting time into higher-value advisory work.

So yes, some tasks will be automated. Some hiring will pause.

The leadership takeaway is simple: if you want AI to improve jobs, you need to deliberately reinvest gains, redesign roles, and build skills. If you do not, inequality inside the firm usually widens, and morale follows it out the door.

A practical AI growth playbook for leaders

Choose your growth bet

Pick one. Serve more customers, improve conversion, launch a new tier, or enter a niche.

Find your constraint

Is it lead flow, onboarding, delivery throughput, quality, or retention. Use AI to remove that constraint, not to chase shiny features.

Ringfence reinvestment

Decide now what you will do with saved time or margin. This is the decision that turns productivity into growth, and growth into hiring.

Redesign roles deliberately

Write down what stops, what increases, and what new work appears. Upskill your team so they control the tools, not the other way round.

Measure outcomes tied to growth

Time-to-deliver, customer satisfaction, throughput, churn, repeat purchase. When these move, headcount decisions become clearer and less emotional.

Conclusion

AI does not decide the future of jobs on its own. Leaders do.

If you use AI to create capacity and reinvest it into better service, new offers, and stronger delivery, you are more likely to see growth and hiring follow. If you use it mainly to cut costs without redesigning the work, you will probably get short-term savings, and long-term headaches, including quality issues, lower trust, and a bigger skills gap inside the business.

The fairest outcome is not automatic either. The organisations that win tend to be the ones that invest in people, not just tools.