South Africa needs more digital talent, but the latest research suggests that producing more graduates is only part of the solution. The greater challenge is preparing people for the work employers actually need done.
South Africa has a difficult contradiction to explain. Employers continue to struggle to recruit skilled technology professionals, yet many graduates struggle to secure their first meaningful employment opportunity. Artificial intelligence (AI) is creating new opportunities, but it is also changing the skills employers expect from people entering the workforce.
The 2026 IITPSA ICT Skills Survey and Research Report, prepared by Africa Analysis, provides important evidence for understanding this contradiction. Its findings suggest that South Africa’s ICT skills gap is not simply a shortage of qualifications. It is also a shortage of practical experience, industry-aligned capabilities and opportunities to develop those skills.
According to the survey, 64.8% of responding organisations experience a moderate or high operational impact from ICT skills shortages. Meanwhile, 61% of ICT practitioner respondents identify AI and machine learning as areas of significant skills shortage.
These are important findings, but percentages alone tell us very little about what should happen next.
To understand the ICT skills gap, we need to examine where the shortages originate, how AI is changing recruitment, why technology skills increasingly depend on human judgement, and what employers, educators and individuals can do differently.
The skills shortage is also a workplace-readiness problem
The survey identifies lack of workplace experience as the leading perceived cause of South Africa’s ICT skills shortage, accounting for 24% of corporate responses. Shortcomings in education and training follow at 20%.
Combined, these two factors account for 44% of employers’ reported primary causes of the skills gap.
This does not mean that 44% of South African graduates are inadequately trained. It means that, among the surveyed organisations, nearly half of the primary causes identified relate directly to the transition between acquiring knowledge and applying it in the workplace.
Consider a graduate who understands programming principles, database design and cybersecurity concepts. An employer might nevertheless require someone who can work with existing systems, diagnose problems, manage deadlines, collaborate with other departments and recognise when a technical solution creates a business risk.
The graduate may possess the theoretical foundation but lack the experience needed to perform independently.
That distinction matters because a qualification demonstrates what someone has studied; workplace readiness demonstrates what they can apply.
Why employers say the skills gap persists
| Reported primary cause | Share | What it suggests |
|---|---|---|
| Lack of workplace experience | 24% | More practical exposure and work-integrated learning are needed |
| Education and training system gaps | 20% | Curricula and training must respond more effectively to workplace needs |
| Lack of diversity | 16% | The talent pipeline may be excluding potential contributors |
| Skills migration | 12% | Retaining experienced professionals is part of the solution |
| Supply-demand mismatch | 12% | Available skills do not always match the roles employers need to fill |
| Rapid technological change | 8% | Skills can become outdated faster than training systems adapt |
| Other factors | 8% | The shortage has additional causes not captured by these categories |
Source: 2026 IITPSA ICT Skills Survey, Exhibit 6.7, page 33.
Figure 1: Why employers say the ICT skills gap persists
The findings raise a question for employers as much as they do for educational institutions: if organisations need experienced professionals, who is responsible for creating that experience?
Graduates cannot acquire three years of relevant workplace experience without access to meaningful work. Where entry-level positions disappear or employers expect junior candidates to perform at intermediate levels, the shortage of experienced professionals risks becoming self-perpetuating.
One response is to expand paid internships, structured graduate programmes, supervised projects and work-integrated learning. Employers can also redesign entry-level roles so that new employees develop practical capability under experienced mentors rather than being expected to arrive fully trained.
For educational institutions, practical projects should demonstrate more than technical knowledge. Students should have opportunities to explain business requirements, manage projects, work in teams, test solutions and present results.
The measure of success should extend beyond how many students complete a qualification to how effectively they can progress into meaningful work.
AI has become a priority, but that does not mean every job is becoming an AI job
One of the clearest developments in the survey is the shift in corporate technology priorities.
In 2022, 12% of responding organisations identified AI and machine learning as a technology priority. By 2026, the figure had reached 34%.
That is an increase of 22 percentage points. In relative terms, the share identifying AI as a priority was approximately 2.8 times its 2022 level.
How corporate technology priorities changed
| Technology | 2022 | 2026 | Change |
|---|---|---|---|
| AI and machine learning | 12.0% | 34.0% | +22.0 percentage points |
| Information security | 24.0% | 28.3% | +4.3 percentage points |
| Data management and storage | 8.0% | 20.8% | +12.8 percentage points |
| DevOps | 5.9% | 18.6% | +12.7 percentage points |
| Big data and data science | 19.0% | 17.0% | -2.0 percentage points |
Source: 2026 IITPSA ICT Skills Survey, Exhibit 6.4, page 31. These are comparisons between survey waves, not necessarily the same organisations.
The most revealing development is not simply that AI has moved to the top. Data management, DevOps and information security remain important.
This is significant because AI depends on capabilities that organisations cannot afford to overlook.
An organisation may want to introduce predictive analytics, automate customer interactions or use generative AI to support business decisions. But an AI system trained on unreliable information may produce unreliable outputs. Weak cybersecurity controls can expose sensitive information. Poorly integrated systems can prevent AI tools from being deployed effectively.
AI adoption therefore creates additional demand for the people who manage data, build software, maintain infrastructure and protect information.
For professionals considering careers in technology, the lesson is to avoid viewing AI as an isolated specialisation. Skills in programming, data management, cybersecurity, statistics and system design can provide important foundations for working with AI technologies.
For employers, it means assessing the capabilities required to make AI useful before purchasing tools or recruiting specialist talent.
A company with unreliable operational data may gain more from strengthening its data management capability than immediately hiring an AI engineer.
The demand for AI skills is growing beyond specialist positions
The IITPSA report also draws on Pnet’s 2025 job market research, which found that demand for general AI proficiency in advertised positions had increased by 488% since 2019.
That is a substantial increase, but the figure is easily misunderstood. The more useful question is how employers distinguish specialist AI expertise from general AI literacy.
The IITPSA corporate survey provides some direction.
How employers evaluate AI capability during recruitment
| Importance of AI capability | Employer responses | Practical interpretation |
|---|---|---|
| Critical requirement | 9.1% | AI capability is considered essential to recruitment |
| Important requirement | 13.6% | AI competence carries significant weight |
| Desirable, but not essential | 54.5% | AI skills may strengthen a candidate’s application |
| Not required | 22.7% | AI capability is not currently considered necessary |
Source: 2026 IITPSA ICT Skills Survey, Exhibit 6.15, page 42.
The combined result is revealing. Among surveyed organisations, 77.2% assign at least some value to AI capability during recruitment, but only 22.7% describe it as important or critical.
This suggests that AI literacy is becoming a competitive advantage without being a universal employment requirement.
A financial analyst may benefit from understanding AI-assisted modelling. A marketing professional may use AI for research and audience analysis. An HR practitioner may need to evaluate AI-supported recruitment tools, including their potential for bias. A public-sector manager may need to understand the implications of using automated systems in service delivery.
None of these professionals necessarily needs to become a machine learning engineer. They do, however, need to understand what AI can do, where it can fail and when human intervention is required.
This creates an opportunity for education providers and employers to distinguish between foundational AI literacy, applied AI skills and specialist technical training rather than treating all AI education as interchangeable.
Is AI replacing workers, or changing the work they do?
The survey’s findings on AI and employment deserve particular attention because they challenge both uncritical optimism and predictions of immediate mass job losses.
Among ICT practitioner respondents, 42% said AI had reduced the complexity of their work, while another 37.6% said it had reduced their workload.
Together, 79.6% reported one of these two benefits.
Yet 8.3% reported increased workloads and 4.4% reported increased task complexity. Combined, 12.7% experienced one of these negative effects.
Figure 3: How AI is changing practitioners’ day-to-day work
The distinction between workload and task complexity is important.
A developer who uses AI to identify possible coding errors may still work the same number of hours but spend less time on difficult troubleshooting. A business analyst who automates routine reporting may have fewer repetitive tasks and more time for complex analysis.
In both cases, AI changes the nature of work without necessarily eliminating the position.
However, these figures describe practitioners’ reported experiences. They are not independently measured productivity gains, and they do not tell us whether organisations have increased or reduced overall employment.
That limitation becomes especially relevant when considering job security.
The survey finds that 45.3% of practitioners perceive AI as posing a low risk to their current roles, while 21.4% see no risk. At the other end, 27.4% perceive moderate risk and 6% perceive high risk.
In other words, approximately two-thirds see little or no immediate personal threat, while roughly one-third recognise a moderate or high level of exposure.
The fact that practitioners currently feel relatively secure does not establish what will happen to employment over the next five years.
It tells us how the respondents perceive their positions at the time of the survey.
What is really motivating employers to adopt AI?
Another finding deserves closer examination. Only 2% of responding organisations identified talent shortages as the primary driver of AI adoption. By comparison, 35.3% identified productivity gains, 29.4% competitive advantage, 17.6% cost efficiency and 15.7% scalability.
This matters because organisations experiencing skills shortages may also be enthusiastic AI adopters, but the two developments should not automatically be treated as cause and effect.
Employers appear to be adopting AI primarily because they expect improvements in business performance.
Whether those improvements translate into more employment, fewer roles or different kinds of jobs will depend on how organisations redesign work and invest in their people.
For managers, the practical implication is to measure AI projects against specific outcomes. Did a tool reduce processing time? Improve accuracy? Free employees to perform higher-value work? Create new risks requiring human oversight?
These questions provide a more useful basis for workforce decisions than assuming that AI adoption is automatically equivalent to job replacement.
South Africa’s digital skills problem begins long before university
There is another reason the ICT skills shortage cannot be addressed through short courses alone.
The supply of future technology professionals depends on the quality of the educational foundations available to learners years before they enter higher education.
According to Department of Basic Education figures cited in the IITPSA report, the number of matric candidates writing Mathematics declined from 269,734 in 2022 to 251,488 in 2024.
That represents 18,246 fewer candidates, or a decline of approximately 6.8% over the two-year period.
The consequence is not that every learner who avoids Mathematics is excluded from the digital economy. Many digital careers involve different levels of mathematical knowledge.
However, advanced areas such as machine learning, engineering, data science and certain software development disciplines depend heavily on mathematical and analytical foundations.
When fewer learners pursue these subjects, the potential pipeline into some technically demanding qualifications becomes more constrained.
Infrastructure creates a further challenge.
The report cites 2024 public-school infrastructure data showing that only 21.8% of public schools had internet connectivity available for teaching and learning.
This is substantially different from simply having an internet connection for administrative purposes.
A school may be able to submit reports online without giving learners regular access to digital learning tools.
The distinction matters because meaningful digital education requires opportunities to use technology, not merely the existence of connectivity somewhere on school premises.
Strengthening the future ICT workforce therefore requires a long-term approach that includes foundational literacy and numeracy, teacher development, mathematics and science participation, reliable connectivity and practical digital learning.
University and workplace interventions remain important, but they cannot completely compensate for educational inequalities that begin much earlier.
The global market for South African talent is already here
South Africa’s employers are no longer competing exclusively with organisations operating within the country’s borders.
The IITPSA practitioner survey finds that 22.2% of respondents are already working remotely for international organisations. A further 33.3% are planning to do so, while 30.8% are considering the possibility.
Only 13.7% reported no plans to pursue international remote work.
These figures should not be interpreted as a prediction that the majority of ICT professionals will leave South African employment.
Planning or considering international work does not guarantee that an individual will secure such an opportunity.
Nevertheless, the findings reveal how international employment opportunities are influencing career choices.
A skilled software developer can potentially work for a company in another country without relocating. That creates opportunities to earn internationally while remaining in South Africa, but it also expands the pool of employers competing for experienced local talent.
This has implications for retention.
In the corporate survey, 44% of respondents reported flexible working arrangements as a retention strategy, while 41% offered development programmes. Performance bonuses were reported by 39%, and increased basic pay by 36%.
These are measures organisations say they use, not evidence that one strategy is more effective than another.
Still, the prominence of development opportunities suggests that employers recognise the importance of professional growth alongside remuneration.
Companies unable to match international salaries may need to compete through meaningful work, learning opportunities, exposure to modern technologies, career progression and supportive workplace cultures.
A retention strategy built only around annual salary adjustments may overlook why highly skilled people are attracted to other opportunities.
The future of learning is not degrees versus short courses
One of the more useful findings concerns how existing ICT professionals have developed their skills.
Formal education remains important. Some 74.5% of practitioner respondents identified degree or diploma programmes as a source of skills acquisition.
At the same time, 55.2% cited on-the-job experience and 45.5% identified professional or vendor certifications.
These responses are not mutually exclusive. A professional may acquire skills through all three pathways.
Figure 4: Where ICT practitioners have acquired their skills
The figures suggest that formal education and continuous professional development should complement one another.
A degree may provide grounding in theory, analytical thinking, computer science and problem-solving. Workplace experience allows that knowledge to be applied under real business conditions. Certifications and shorter programmes can help professionals respond to technologies that have changed since they completed their formal studies.
The survey also finds that 66.2% of practitioners prefer online self-paced learning, compared with 35.9% who selected academic programmes as a preferred learning format.
These preferences do not imply that academic qualifications are losing their value. They reflect the flexibility working professionals increasingly need to develop new skills while managing existing responsibilities.
This combination is already visible across different forms of education.
For example, the Regenesys School of Technology offers formal qualifications such as the Bachelor of Science in Computer Science and Postgraduate Diploma in Data Science.
The Regenesys School of AI addresses different levels of AI capability through its Leadership, Managerial, Tactical and Masterclass series, including areas such as responsible AI governance, organisational transformation and practical applications.
Meanwhile, Digital Regenesys offers skills-focused courses including Artificial Intelligence, Data Science with AI, Cybersecurity with AI, DevOps and Cloud Computing with AI, Data Analytics and AI Leadership.
Workplace-based development forms another part of the picture, with providers such as Regenesys Corporate Education offering customised management, leadership and organisational learning programmes.
These pathways address different needs. No single qualification, certificate or training provider can resolve the skills shortage independently.
The challenge is ensuring that learning opportunities connect to demonstrable capability, workplace application and continued professional development.
What should individuals, employers and educators do with these findings?
The value of the IITPSA survey lies in its ability to inform decisions.
For someone entering the workforce, the findings suggest that technical qualifications should be supported by evidence of practical application. A portfolio demonstrating the ability to solve real problems may help employers understand what a candidate can contribute beyond their academic results.
For experienced professionals, the priority is not necessarily to abandon an established career and pursue a specialist AI qualification. It may be more valuable to identify where AI, data analysis or automation can strengthen existing expertise.
For organisations, buying technology without developing the people expected to use it risks creating an expensive implementation problem rather than a productivity improvement.
Turning the findings into action
| Who | What the evidence suggests | A practical next step |
|---|---|---|
| Students and graduates | Employers value applied experience | Build a portfolio with a completed project, documented process and measurable outcome |
| Working professionals | AI literacy is increasingly desirable | Identify one relevant AI skill and demonstrate its practical use in your current work |
| Employers | Skills shortages affect operations | Conduct a skills audit and introduce structured mentoring, internships or graduate development |
| Senior leaders | AI adoption is driven by business outcomes | Evaluate projects against productivity, quality, risk and workforce development measures |
| Education providers | Qualifications need stronger workplace alignment | Co-design projects with employers and expand opportunities for practical learning |
| Policymakers | The skills pipeline begins at school | Prioritise foundational STEM education and connectivity that enables teaching and learning |
These recommendations are not instant solutions. Building digital capability requires sustained investment, and individuals face different financial, educational and workplace circumstances.
But the evidence suggests that greater coordination between education and employment would be more useful than treating training and recruitment as separate problems.
We also need better ways to measure progress
South Africa’s ICT skills shortage is frequently discussed through the number of vacancies, the demand for specific technologies and the qualifications available to prospective students.
These indicators matter, but they do not reveal whether the country is becoming better at developing and retaining capable professionals.
A stronger skills-development scorecard would examine whether graduates secure relevant employment, whether workplace programmes lead to sustained career progression, whether employees can demonstrate newly acquired capabilities and whether businesses achieve measurable improvements from digital investment.
It would also examine who remains excluded from opportunities to develop digital skills.
Without these measures, the country risks repeating the same diagnosis year after year while struggling to establish which interventions are working.
The 2026 IITPSA survey provides a useful starting point because it connects employers’ concerns with practitioners’ experiences and the wider educational pipeline.
However, the survey reflects the views of participating organisations and ICT professionals. Its corporate and practitioner percentages refer to different respondent groups, and the report does not clearly disclose the overall sample sizes and response rates. The findings should therefore be treated as evidence of reported trends rather than a national count of vacancies or a prediction of employment outcomes.
South Africa needs a stronger skills pipeline, not simply more AI courses
AI will continue to influence the types of capabilities organisations need. The IITPSA findings indicate that this change is already visible in technology priorities, recruitment expectations and the daily experiences of ICT professionals.
Yet the report’s most important message extends beyond AI.
South Africa needs stronger foundations in education, more opportunities for graduates to acquire practical experience, continuous learning for established professionals and better conditions for retaining scarce talent.
The country’s response should not be measured simply by how many people enrol in AI programmes or how many organisations introduce AI tools.
It should be measured by how effectively people can use what they learn, how successfully organisations develop their talent and whether technological progress creates meaningful economic opportunities.
The real test is not whether South Africans can learn to use AI. It is whether the country can turn expanding digital knowledge into a workforce capable of solving real problems.
Research basis: 2026 IITPSA ICT Skills Survey and Research Report, prepared by Africa Analysis and issued October 2026. All survey percentages refer to respondents, unless otherwise stated. Historical education data and Pnet job-market figures are identified separately.
