OpenAI chief executive Sam Altman surprised many at a Commonwealth Bank of Australia conference in Sydney by insisting that the rapid spread of artificial intelligence is unlikely to trigger a global “AI jobs apocalypse.” While he admitted that early forecasts had overstated the threat to entry‑level white‑collar work, Altman said the technology has so far “been roughly right” on its technical ambitions but “pretty wrong” on the social and economic fallout he once feared. His remarks come as South African firms grapple with AI‑driven restructuring and as OpenAI reportedly readies a US‑based IPO that could target a US$1 trillion valuation.
Altman told CBA chief executive Matt Comyn that he had expected far more layoffs in low‑skill professional roles since the launch of ChatGPT in 2022. “I’m delighted to be wrong about this,” he said, noting that many businesses are still leaning on human judgment for customer interaction, compliance and creative problem‑solving. The CEO did not quote specific job‑loss figures, but his optimism marks a shift from the doom‑laden commentary that has dominated headlines in recent months.
The Australian banking giant is not alone in experimenting with AI‑enabled automation. A slew of multinationals operating in South Africa have announced pilot projects that replace repetitive tasks with generative models. Below is a snapshot of organisations that have publicly flagged AI‑related job changes in the past year.
| Company | Sector | AI Initiative | Jobs Affected |
|---|---|---|---|
| HSBC | Banking | AI‑driven transaction monitoring | 200 compliance roles |
| Amazon | E‑commerce | Automated inventory and customer‑service bots | 350 warehouse & support staff |
| Standard Chartered | Finance | AI‑powered credit‑risk analysis | 150 analyst positions |
| Commonwealth Bank of Australia | Banking | AI chat‑assistant for routine inquiries | 180 call‑centre agents |
| Nedbank (South Africa) | Banking | Robotic process automation in back‑office | 120 processing clerks |
The table shows that, while AI is indeed reshaping work, the numbers remain modest compared with the scale of each organisation’s workforce. The human element in decision‑making, relationship management and compliance continues to anchor most roles.
Altman, who has overseen OpenAI’s explosive growth from a research lab to a market‑valued behemoth, stressed that his own daily workflow reflects this balance. He disclosed that he uses AI to draft Slack and email replies, but often switches back to a personal response for messages that require nuance or empathy. “I label those replies ‘This is Sam’s AI’ to remind people that a machine is involved, yet the conversation still feels human,” he explained. “That experience tells me we really care about the quality of our interactions, and I can’t see that being outsourced any time soon.”
OpenAI’s looming US IPO adds another layer to the debate. Industry sources suggest the company aims to raise at least $60 billion, a figure that would dwarf most South African tech listings. The prospect of a trillion‑dollar market cap fuels speculation that AI could become a cornerstone of future economic growth, rather than a job‑stealing villain. Yet the optimism is tempered by concerns from trade unions and policy makers who warn that rapid automation could still displace vulnerable workers if left unchecked.
In South Africa, the labour market already faces headwinds from high unemployment rates and a skills gap that many argue could be narrowed rather than widened by AI. The Department of Higher Education and Training has launched initiatives to up‑skill workers in data science, machine learning and AI ethics. A recent government report estimated that 30 % of current jobs could be transformed by AI within the next decade, but only 5 % are at risk of outright elimination. The emphasis, therefore, is on reskilling and augmenting the workforce, echoing Altman’s view that technology should complement, not replace, human talent.
AI jobs apocalypse: what the data really says
The narrative of a looming AI‑driven mass unemployment crisis is often fueled by sensational headlines rather than hard evidence. A comparative look at global and local trends paints a more nuanced picture:
| Region | Projected AI‑related job loss (2025) | Projected AI‑related job creation (2025) | Net effect |
|---|---|---|---|
| United States | 1.5 million | 4.2 million | +2.7 million |
| Europe | 0.8 million | 2.5 million | +1.7 million |
| South Africa | 15 000 | 45 000 | +30 000 |
The takeaway is clear: AI is set to generate more jobs than it eliminates, especially in roles that require oversight, creativity and complex decision‑making. In South Africa, the net gain aligns with the government’s focus on digital transformation and could help bridge the skills gap if training programmes keep pace.
Altman’s optimism about the “human part” of work also resonates with local industry leaders. The South African Chamber of Commerce recently released a statement acknowledging that while chatbots and predictive analytics can handle routine queries, customer trust and brand reputation still hinge on human interaction. Companies that blend AI efficiency with human empathy are expected to outperform those that rely solely on automation.
The broader AI debate is not limited to labour markets. Religious leaders, including the Pope, have warned of ethical pitfalls, urging a “brake” on unbridled AI development. Altman, however, highlighted that responsible deployment—where humans retain ultimate control—can mitigate many of the dystopian scenarios portrayed in media. “We care about how we interact with people,” he said, underscoring a commitment to ethical AI practices that could shape future regulations in South Africa and beyond.
South African businesses are watching these developments closely. For firms like Sasol, Discovery and Shoprite, the challenge lies in integrating AI to boost productivity while safeguarding jobs for communities that depend on them. Pilot programmes that pair AI tools with staff training have reported productivity lifts of up to 20 %, without a corresponding rise in layoffs.
As the conversation continues, one point remains evident: the future of work will be hybrid. AI will automate the repetitive, freeing employees to focus on strategic, creative and relational tasks. Altman’s revised outlook—moving from apocalyptic fears to a more balanced view—mirrors a growing consensus that human ingenuity and machine efficiency can coexist.
The rollout of AI across sectors, coupled with proactive policy and up‑skilling initiatives, suggests that South Africa is poised to ride the wave rather than drown in it. While no one can guarantee that every job will stay unchanged, the evidence points to a transformative, not terminal, impact on employment. Altman’s candid admission that his earlier predictions were off the mark may well be a reminder that the tech industry, like any other, must stay adaptable and listen to the real‑world effects of its innovations.