AI adoption in the workplace increasing, but job fears and shadow AI remain concerns.
Organisations across the Global South are rapidly integrating AI into the workplace, with widespread adoption reported even as concerns about job displacement intensify, particularly among frontline workers.
This is according to the third edition of Boston Consulting Group’s (BCG) annual AI at Work 2025: Momentum Builds, But Gaps Remain survey (based on responses from over 10 600 employees in 11 countries, including 509 from SA), which reveals that 72% of South African employees use AI at least several times a week – matching the global average. India (92%) and the Middle East (87%) lead in regular usage.
“These countries are leapfrogging traditional development stages, using AI to tackle systemic challenges in healthcare, education and agriculture,” the report notes.
Despite high adoption, apprehension remains. Globally, 41% of employees fear automation could eliminate their jobs within the next decade. Managers and leaders (43%) are more concerned than frontline workers (36%). In SA, however, the trend is reversed: 38% of frontline employees fear job loss – on par with the global average – while only 24% of local managers share the same concern, among the lowest globally.
Jacqueline Foster-Mutungu, MD and partner at BCG Johannesburg, likened the AI shift to the continent’s mobile technology leap: “This strong adoption of AI is, in a sense, similar to Africa’s leapfrogging of traditional landline infrastructure.”
The report also attributes the rapid uptake to relatively light AI regulation in many Global South regions, enabling faster deployment compared to stricter environments like the European Union.
Sylvain Duranton, global leader of BCG X, said among the key takeaways from the research is that proper training, leadership support and access to the right tools can address many of the challenges associated with AI adoption and application.
A separate global study by Salesforce’s Slack Workforce Index shows daily AI use among desk workers has surged by 233% in the past six months. The survey of 5 000 workers found that daily AI users are 64% more productive and 81% more satisfied with their jobs than non-users.
“AI adoption is accelerating rapidly, and it’s changing the way the world works for the better,” said Peter Doolan, EVP and chief customer officer at Slack. “Businesses are unlocking greater productivity and enabling employees to focus on work that drives real impact and growth.”
The study also reveals a shift in AI usage: 96% of respondents report using AI to complete tasks they previously lacked the skills for. Workers are now 154% more likely to use AI to enhance quality and creativity, rather than just efficiency.
“As workers actually use and experiment with AI agents, their trust and enthusiasm grow,” added Lucas Puente, VP of Research at Salesforce.
BCG’s report cautions that real transformation requires more than just deploying AI tools. True value is being realised by a small subset of companies that redesign workflows and invest in upskilling.
“Companies cannot simply roll-out GenAI tools and expect transformation,” said Duranton. “Returns come when businesses invest in people, rethink how work gets done and align leadership around an AI strategy.”
AI agents – the next phase
Three-quarters of global employees believe AI agents – intelligent digital assistants capable of autonomous task management – will be crucial to future success. Yet, only 13% say these tools are currently integrated into their workflows. SA mirrors this global average.
Countries leading in AI agent adoption include Brazil (18%), India (17%), Spain (16%) and the US (15%). Only one in three employees globally understand how AI agents function.
However, familiarity appears to reduce fear. As workers become more accustomed to AI agents, they begin to see them as collaborators rather than threats.
Despite this, AI agents are not widely deployed. In practice, only 13% see agents integrated into broader networks.
At the same time, organisations are warned to monitor the emergence of shadow AI – the unauthorised use of AI technology that deliberately bypasses IT and security mechanisms.
“Shadow AI comes with a lot of risk because usually it means people have not been properly trained. They have not been educated on responsible AI. The way to prevent it – and the mitigation strategy – can only be about training people, and actually be more bullish about the tools that you give access to,” said Duranton.
