If you’ve looked at your phone this week you’ve most likely come across another article cautioning that artificial intelligence is heading towards your job, your inbox, and possibly even your peace of mind if the algorithm behaves badly. Since WhatsApp groups are full of concern and boardrooms are occupied by discussions about ‘AI strategies’, it’s reasonable to ask yourself whether your job is safe from AI.
The short answer is that certain jobs are safe whereas others aren’t, and most of them will undergo some change. The longer answer is where things become interesting and even empowering.
Begin by considering the fear—this is not unreasonable. The latest development in generative AI differs from earlier forms of automation. Rather than simply speeding up tasks such as moving boxes or calculating numbers more accurately, it can now replicate aspects of creative, administrative, and analytical work which have always been regarded as “uniquely human”. What used to take hours can now be done in just minutes, for example, drafting a memo, summarising a meeting, or turning data into a slide. Given that in South Africa we are already dealing with high unemployment, and there are areas with limited digital access, the situation is serious.
—but there’s a difference in that AI tends to disrupt tasks rather than entire jobs. That point is significant. A 2023 study by the International Labour Organisation discovered that although generative AI is likely to automate some clerical and routine tasks, most jobs are more likely to be altered than entirely replaced. To put it simply, AI affects the way you work rather than merely affecting whether you work. It’s similar to having a highly efficient colleague who never sleeps, occasionally invents things, and still requires someone to keep an eye on them.
How might it look in the various sectors?
- Administration and support: AI performs well in areas such as scheduling, preparing meeting minutes, capturing invoices, and planning travel. If your role involves a lot of paperwork, you should anticipate that things will change. The greatest risk is associated with jobs that are mainly task-oriented and involve little interaction with people. However, for those of you who keep everything going by coordinating teams, understanding the overall situation, and building relationships, there is still plenty of opportunity. The best approach is to become the person who knows how to use these new tools.
- Marketing and media: AI can write copy, suggest headlines, and make rough video edits. It’s good at first drafts, but it struggles with subtlety. In South Africa’s diverse market, cultural understanding, switching between languages, and telling stories that fit the context still make the difference between something memorable and something forgettable. Editors and strategists who can guide AI, check its work, and add local insight will do well.
- Law and Finance :In the fields of law, finance, and consulting, activities such as reviewing documents, carrying out due diligence, and summarising policies can all be accelerated by the use of AI. That said, it is still people who should deal with judgement, build up client trust, and understand the regulations. You can anticipate a reduction in the number of billable hours per project, but there will be greater demands placed on strategic thinking. The kind of professionals who will continue to be essential are those who concentrate on higher-level work rather than competing with the software for routine tasks.
- Healthcare and education: In the fields of healthcare and education, AI provides assistance rather than taking over in clinics and classrooms. Empathy, ethical considerations, and the practical complexities involved are important. AI can be used to draft summaries of patient care, prepare lesson plans, and identify potential risks. However, it is the relationship that constitutes the key aspect of the work. The danger is not so much about jobs being lost as it is about task inflation: there will be more administrative work, more dashboards, and less time available. It will be the decisions made by leaders that decide whether AI adds back hours or takes them away.
- Regarding technology and data: It’s interesting that AI is now producing code which used to be carried out by junior developers. This situation may make it more difficult to begin a career at the entry level. At the same time, there is a rapidly increasing demand for individuals who can integrate systems, think about products, design prompts, manage data, and deal with cybersecurity. If you are working in the tech field, you should expect to have to continue learning as part of your job.
- Retail and customer service: Chat bots will be dealing with a greater number of customer queries, but South African consumers usually prefer to speak to a real person when difficulties arise. The factor that will distinguish one company from another is the ease with which they transfer customers to human agents, the degree of empathy they show, and their ability to resolve issues. Employees on the shop floor are moving away from scanning bar codes and are instead concentrating on providing advice and assisting customers with making their choices.
There’s a clear pattern: the kinds of jobs that involve human judgement, an understanding of context, interpersonal skills, and expertise are the most secure, while those which are mainly predictable and deal with documents are at the greatest risk, unless the individual carrying out these tasks learns to steer, verify, and explain the AI.
Well, what does this mean for your actual Monday?
Check over the tasks you carry out.
Not your job title—your tasks. Make a list of the things you do in a typical week and place a mark alongside them:
- Involving repeated actions following a set of rules (for example, preparing similar emails or compiling reports)
- Variable and context-heavy. For example, negotiating timelines or resolving conflicts.
- Innovative and strategic (for example, in the development of campaigns and advising clients)
The first bucket is ripe for automation. Your goal is to move time from bucket one to buckets two and three and to learn how to coordinate bucket one using AI.
Learn the tools properly since merely providing a prompt does not amount to mastery.
This month, spend ten hours on practical, hands-on practice:
- Draft a prompt library for your everyday tasks
- Learn to ask for sources, formats, and constraints
- Build a review checklist to catch AI errors
- Try one work flow end-to-end: brief, generate, verify, deliver
Make certain that you understand the company’s data policy if one exists, or ask for one to be set up if it doesn’t. You should never paste sensitive information into public tools; to claim that this is paranoid is incorrect; what is correct is to deal with the situation professionally.
Claim the human layer.
Make yourself the person who can:
- Translate messy needs into structured briefs
- Check for bias, accuracy, and tone
- Communicate decisions clearly and ethically
- Hold the client or team relationship
These are not “soft” skills. They are rare skills.
Assess your worth.
Since AI is taking over routine jobs, some managers are going to ask why they still need such a large number of people. In order to prepare for this line of reasoning, present them with the results:
- Faster turnaround times
- Higher client satisfaction
- Fewer errors
- New initiatives launched because you freed up time
Keep a basic evidence log. Data saves jobs.
Up skill with purpose.
Not all certificates are equal. Prioritise:
- Data literacy: reading dashboards, spotting dodgy stats
- AI fluency: prompts, verification, workflow design
- Domain depth: the substance of your industry
- Communication: writing, presenting, facilitation. In South Africa, because connectivity
costs and access vary, consider courses that use little data, employer-organised learning programmes, or community workshops. Combine online resources with local networks like women’s discussion groups, professional associations, and alums groups.
Keep an eye on the ethics and on your own rights.
Since AI poses a number of risks—such as biased decisions, productivity monitoring similar to surveillance, and vague boundaries concerning intellectual property—it’s necessary to find out how AI tools are tested, what data they collect, and how their performance is evaluated. It is important to have a collective voice, so raise your concerns in internal forums or in unions if suitable. The question of how AI should be introduced responsibly should be decided by leadership, not by IT.
Alright, let’s consider the present situation in a realistic way. Some jobs are going to be cut because those which require a lot of clerical work are facing pressure as software becomes better at extracting, summarising, and storing information. In the case of individuals in such vulnerable positions, the right human response is not to advise them to ‘learn to code’, but instead to offer them specific reskilling opportunities, provide bridge financing, and ensure that equitable pathways are available for them to move into new roles. Companies that introduce AI should also promise to upskill their existing employees and reassign staff, not just reduce their workforce. Furthermore, government policy should play a part by giving training incentives, establishing transparency standards, and putting in place safeguards against discriminatory systems.
There’s also a hidden opportunity. Many South African teams are burdened by administrative work and don’t have sufficient time for strategy; AI—when used properly—can free up hours, which can then be put towards mentoring junior staff, improving products, visiting clients or thinking creatively. The true value of time is seen when leaders reorganise their work rather than simply assigning more tasks. In order for this to take place, leaders have to have both the courage and clarity to know what they really want to achieve.
A final word on mindset.
Fear causes the mind to become narrow, whereas curiosity has the effect of broadening it. The workers who are successful will not see AI as merely a temporary fad, nor will they regard it as something that is predestined. On the contrary, they will carry out tests, take measurements, and make adjustments. They will be generous in sharing their findings. They will remember that technology is a tool and that the form which tools take is determined by the hands that use them.
As for whether your job is protected against AI, safety is not a fixed state but one that must be actively worked towards. It requires learning a small amount every week, posing better questions, and slowly progressing towards the parts of your job that are entirely within your own capability.
What this signifies is the quiet revolution that is presently taking place; it has nothing to do with robots taking over, but instead concerns people’s development in the areas of judgement, relationships, and purpose. In a nation which has a well-established record of resilience and reinvention, this is a story which we can all help to write together.
The International Labour Organisation’s 2023 report, entitled ‘Generative AI and Jobs: A global analysis of potential effects on job quantity and quality’, states that most jobs are more likely to be augmented than automated, although administrative positions face the greatest risk of automation. The report also stresses the need for skills development and policy responses.
