Artificial intelligence promises greater efficiency, but it could also close off opportunities for the very people who need them most. For HR leaders, the challenge is to ensure that technological progress creates a more inclusive workforce rather than reinforcing existing inequalities. How organisations handle AI and inequality now will shape who gets access to work in future.
Artificial intelligence (AI) is rapidly changing how organisations work, recruit, develop talent and make decisions. The business case for greater efficiency is compelling, particularly when resources are stretched and expectations continue to rise.
Yet there is a question that deserves considerably more attention in executive discussions: who benefits from these efficiencies, and who risks being left behind?
For Leanne Emery Hunter, CEO of Tshikululu Social Investments, this question extends beyond technology adoption to the responsibilities of business leaders in a country facing profound structural inequality.
Speaking to Alan Hosking on the HR Future Podcast, Hunter highlighted the enormous potential of AI to improve education, healthcare, economic participation and social development. She also warned that without deliberate intervention, these advances could reinforce the very inequalities they have the potential to address. That tension between AI and inequality sits at the centre of her argument.
Her observations hold important lessons for CHROs, particularly as organisations reassess workforce structures, entry-level employment, skills development and their broader contribution to society.
The entry-level employment problem HR cannot afford to ignore
- Redesign entry-level roles instead of automatically eliminating them. Identify tasks that can be automated while retaining opportunities for employees to develop judgement, communication and problem-solving skills.
- Create AI-assisted apprenticeships. Allow early-career employees to work alongside experienced professionals, using technology to accelerate learning rather than replace participation.
- Measure access to employment. Track whether automation is reducing graduate placements, learnerships and other entry-level opportunities.
- Build alternative routes into employment. Consider skills-based recruitment, work-integrated learning and partnerships with training providers.
2. AI governance needs to address exclusion, not only efficiency
3. Stop measuring social investment by how much money is spent
4. Social investment and workforce strategy should be working together
5. Collaboration requires more than good intentions
6. Put long-term workforce resilience on the executive agenda
AI and inequality: the leadership decision that will shape AI’s impact
One of Hunter’s most important concerns is that AI could eliminate the first rung of the employment ladder.
Many organisations are introducing automation into administrative, transactional and routine roles. These are often the very positions through which young people gain their first exposure to working life.
Removing these roles may deliver immediate productivity improvements, but it also creates a longer-term workforce problem.
If organisations automate the positions where employees traditionally acquire foundational skills, where will their next generation of experienced professionals come from?
What CHROs can do about AI and inequality
The immediate cost savings from automation should be weighed against the future cost of a shrinking talent pipeline.
Hunter raises another concern: technology can amplify the biases already embedded in human decisions and organisational systems.
For HR, this is particularly relevant when AI influences recruitment, performance management, employee development or workforce planning.
An algorithm trained on historical decisions may reproduce existing patterns of disadvantage. An apparently objective process can therefore produce outcomes that are neither fair nor inclusive.
Governance must extend beyond asking whether a technology works to examining whom it serves and whom it may disadvantage.
CHROs should introduce an inclusion assessment before implementing AI in consequential people decisions.
That assessment should examine whether the technology has been tested across relevant demographic groups, whether its recommendations can be challenged, and who is accountable when an automated decision produces an unfair outcome.
HR should also insist on meaningful human oversight, particularly where decisions affect employment opportunities and career progression.
Hunter’s argument places human judgement at the centre of responsible AI adoption. Technology can improve decision-making, but leadership remains responsible for the consequences.
Another valuable lesson from Hunter’s work concerns the difference between funding activities and achieving meaningful outcomes.
She describes the non-profit starvation cycle, where organisations receiving funding are expected to direct almost all their resources towards programme beneficiaries.
While understandable, this pressure can leave insufficient funding for essential infrastructure, including technology, data systems, training and organisational capability.
The result is a programme that may look financially efficient but lacks the foundations required to deliver sustainable impact.
There is a parallel within corporate HR.
Training budgets are often measured by expenditure, participation and completion rates. Yet these figures reveal relatively little about whether employees have developed useful capabilities or improved their career prospects.
A more effective approach is to measure outcomes rather than activity.
Instead of reporting that 500 employees completed an AI training programme, CHROs could measure how many subsequently applied those skills, moved into new roles or demonstrated improved performance.
This requires investment in measurement systems, management capability and follow-through, not simply more training content.
Hunter argues that South Africa’s social challenges cannot be addressed effectively in isolation.
Education, unemployment, economic participation, community development and gender-based violence frequently intersect.
She describes programmes where economic empowerment initiatives must consider the wider conditions affecting participants’ ability to succeed.
This presents an opportunity for CHROs to rethink the relationship between corporate social investment, skills development and workforce planning.
Consider a company struggling to recruit people with specific technical capabilities while simultaneously funding education or employment programmes in surrounding communities.
Rather than treating these as separate initiatives, the organisation could connect them.
HR could identify future skills requirements, collaborate with education providers and community organisations, and create pathways from training into meaningful employment.
Such programmes should also consider practical barriers to participation, including transport, safety, digital access and the availability of workplace support.
The objective should be a functioning pathway into employment, not merely another programme with impressive participation figures.
Hunter observes that companies increasingly recognise the value of collaboration, but coordinating partnerships remains difficult.
Organisations may share similar social investment objectives while continuing to operate separate programmes, duplicate resources and compete for recognition.
For CHROs, this raises an important possibility: could organisations achieve more by collaborating on workforce challenges they cannot solve independently?
Employers operating in the same region or industry often face common skills shortages, limited training capacity and barriers to youth employment.
A shared training initiative, developed with educational institutions, government and community partners, could offer greater scale and continuity than several disconnected projects.
However, successful collaboration requires clear accountability.
Before committing resources, participating organisations should agree on the outcomes they want to achieve, how progress will be measured, who will coordinate implementation and what happens when individual funding commitments change.
Hunter emphasises the importance of shared purpose and long-term outcomes. These are particularly relevant when organisations want programmes to survive beyond annual budgets or individual corporate sponsors.
Perhaps the most significant implication of Hunter’s insights is the need to reconsider the time horizons used in business decisions.
Organisations understandably operate within annual budgets and shorter financial reporting cycles. Social development and workforce transformation, however, often require sustained investment over several years.
This tension is especially important when weighing up AI and inequality.
A technology investment may demonstrate a positive financial return within months, while its effects on employment access and workforce capability, and the wider link between AI and inequality, become visible much later.
CHROs therefore need to bring longer-term questions about AI and inequality into strategic planning.
What capabilities will the organisation require in five years? How will young people access those opportunities? Which employees risk exclusion as technology advances? What investments today will reduce those risks?
These questions are essential to building a workforce capable of adapting to technological change.
AI presents South African organisations with an opportunity to improve productivity while expanding access to education, employment and economic participation.
But these outcomes will not emerge automatically.
Hunter’s insights suggest that responsible adoption requires deliberate investment in people, stronger governance, collaboration and a willingness to measure success beyond immediate financial returns.
For CHROs, the responsibility for managing AI and inequality is substantial. Decisions made today about automation, recruitment, learning and workforce design will influence who gets to participate in tomorrow’s economy.
The question worth taking into the next executive meeting is whether the organisation’s AI strategy is creating new opportunities for people or simply finding more efficient ways to operate without them.
Leanne Emery Hunter is CEO of Tshikululu Social Investments, a South African organisation specialising in social investment strategy, fund management, programme implementation and impact evaluation. Previously Chief Commercial Officer at the Youth Employment Service (YES), she holds an MBA with distinction from the Gordon Institute of Business Science (GIBS). Her work focuses on meaningful social impact, collaboration, digital inclusion and the role of innovation in addressing structural inequality.
