Companies will soon have AI colleagues working alongside human employees in the workplace, according to research by technology solutions firm Prosus and global start-up and VC intelligence platform Dealroom.
The companies have released a report, ‘The Rise of the Agentic Workforce: How Autonomous AI Agents Will Transform the Workplace’, which examines the emergence of agentic AI – autonomous AI systems with the ability to make independent decisions, the capacity to learn and adapt, and to take ownership of completed tasks.
Prosus asserts that this technology marks the beginning of the third wave of AI development, following predictive AI and GenAI.
“The rise of agentic AI represents a foundational change in how we’ll work with AI technology in the coming years,” said Fabricio Bloisi, CEO of Prosus. “We are past the tipping point of AI agent adoption, and it is now firmly rooted in the workplace. Agentic AI companies are attracting billions of dollars of venture capital investment. At Prosus, we plan to hire the largest AI workforce in the industry this year, with AI agents working across our organisation from HR to customer support.”
The report found that agentic AI is quickly gaining traction, VC funding for agentic AI is ramping up, and markets are heading towards a multi-agent future. Moreover, AI agents will serve as digital co-workers rather than just helpful tools, with coding agents the first to reach product-market fit.
Fully fledged AI employees are months, rather than years, away, Prosus claims.
AI funding
Venture Capital funding for agentic AI is ramping up quickly. It is expected to total 10% of all AI funding rounds in 2025, amounting to US$6.7 billion of investment, with Europe punching above its weight in its proportion of funding.
Prosus says customer service and healthcare platforms are the most funded application areas, while robotic process automation and AI agent builders have outperformed within the category of ‘AI agent enablers’.
The company believes markets are heading towards a multi-agent future, where organisations will be structured differently, and which will reward entirely new skillsets.
According to Prosus, people will set the direction for their teams of agent-based ‘digital colleagues’, who help to carry out tasks. As AI takes on more routine tasks, people will need to evolve their skillsets, understand how to work effectively alongside AI systems and guide AI performance, with a focus on strategic thinking and creative problem-solving.
AI factories
As AI ‘co-workers’ are recruited into markets in Africa, their human counterparts must empower themselves with AI skills – and construct AI factories, says President Ntuli, MD of HPE South Africa.
“The continent’s current AI talent gap is already impacting business outcomes, with many organisations struggling to find the skills needed to successfully pilot and implement new AI initiatives. Bridging this divide will require co-ordinated, long-term investment in education, training and professional development. One promising approach is the establishment of AI hubs; centres of excellence where governments, academia and industry can collaborate to accelerate skills development. These hubs can serve as catalysts for innovation, offering access to global advisory services, technical mentorship and hands-on training.”
Ntuli adds that by nurturing a pipeline of homegrown researchers, data scientists and engineers, Africa can not only meet its domestic AI needs but also position itself as a contributor to the global AI economy.
“Countries across the world are increasingly embracing a sovereign AI strategy by building domestic 'AI factories', national computing platforms or networks of AI hubs powered by local data and tailored to national priorities. Denmark, for example, launched a supercomputer funded by its pharmaceutical success to drive biotech and AI innovation. Italy, Sweden, the UAE and India have also developed similar initiatives, aiming to embed cultural identity and strategic autonomy into AI development,” Ntuli continues.
AI factories can serve both public and private sector needs, enabling organisations to access the compute power necessary for training and deploying advanced AI models, he concludes.
Share