2025 will be the year of AI maturity in the enterprise. The evolution of AI has reached a point where increasingly, AI-powered agents will have the capabilities that make them indispensable to knowledge workers. Agentic AI leads this charge by spearheading the use of artificial intelligence that is increasingly an autonomous partner to business, and not a tool that depends on a user.
Johannesburg, 23 January, 2025 – UiPath (NYSE: PATH), a leading enterprise automation and AI company, predicts that 2025 will be the year of AI maturity and the emergence of Agentic Automation.
Greg Williams, Regional Vice President for South Africa at UiPath, says AI is no longer an optional investment, but is now establishing itself as an indispensable way of working across industries through the adoption of agentic automation.
According to Williams, Agentic Automation harnesses the power of generative AI (GenAI) to give agents the capabilities to analyse unstructured data, recognise patterns, plan actions, and make decisions on their own. Agentic Automation is powered by large language models (LLMs), GenAI, and large action models (LAMs), as well as other advanced AI, to enable reliable, autonomous action.
Williams highlights UiPath’s AI and Automation Trends 2025 report and State of the Automation Professional report 2024, which indicate that AI is already widely deployed in automation projects. In South Africa, PWC says 92% of CIOs anticipate widespread AI implementation by this year.
“AI-powered automation will play an increasingly important role in helping South African companies to boost their competitiveness and efficiency. We believe this year will see a growing number of organisations exploring the opportunities and building the necessary capabilities to harness AI,” he says.
He says one indication of the growing focus on AI and automation in South Africa is the increase in registered users of the UiPath Academy open online training platform. Over 200,000 people have enrolled since its launch in 2017, with over 18,000 enrolling for training last year alone.
UiPath’s latest report shows that 61% of companies worldwide have added more automation workers in the last year. But despite fears of job-losses in labour-intensive economies such as South Africa, the report highlights that 81% of surveyed companies plan to hire more people as automation increases.
UiPath outlines key trends for 2025, and how organisations should prepare to embrace them:
TREND I. The rapid development of AI agents that autonomously understand, plan and operate within complex workflows.
Prediction: Agent-based AI will allow companies to automate complex processes with minimal supervision, increasing productivity and opening up new opportunities for industry-specific automation solutions.
Preparation: Subject at least one process to agent automation. Passivity stifles innovation, so it is worth joining the pioneers in Agentic AI or at least keeping a close eye on their activities.
TREND II. With the development of agent-based artificial intelligence, orchestration is becoming increasingly important.
Prediction: In order to realise the full potential of Agentic AI, businesses will need a dynamic infrastructure that enables humans and robots to collaborate with agents. Additionally, it will need to enable their creation, deployment, as well as monitoring of their activities with transparency and compliance.
Preparation: Create a plan for the introduction and scaling of an Agentic AI environment, taking into account the need for human oversight and the role of RPA robots as the ‘hands and feet’ of new agents.
New business strategies
This year’s changes will not only affect the technology behind AI itself, but also the ways in which it can be used in a business context. Organisations will rethink how they approach problems such as technology debt or the lack of a concrete strategy when implementing AI as both the opportunities for its widespread implementation skyrocket, but the costs of falling behind do so too.
TREND III. Executives are disillusioned with spending millions on AI without the expected results. Companies that implement AI without clear strategies to measure and demonstrate ROI will struggle to justify the short-term costs, but rue the costs in the long run.
Prediction: Companies need to develop procedures to track and measure the impact of AI on key business outcomes such as productivity, cost savings and revenue growth. The role of business technology providers will also increase. One of the most widespread uses of AI are ‘copilots’, which help employees with various office tasks. These are being developed by major enterprise technology providers such as Microsoft, GitHub and Google and are yielding excellent results. For example, UiPath has created Autopilot for Developers, which reduces automation development time by 75%.
Preparation: A greater focus on AI ROI measurement tools and strategies, a thorough analysis of the company’s use of the technology and training in the use of embedded tools such as ‘copilots’. With greater integration opportunities and the increasingly autonomous nature of AI, the ROI-calculation needs to evolve beyond time-savings.
TREND IV. Heads of technical departments are turning to AIOps (Artificial Intelligence for IT Operations) tools to streamline operations, automate routine tasks and improve system reliability.
Prediction: As organisations increasingly rely on complex digital infrastructures, many are struggling with technology debt – the cumulative weight of compounding inefficient processes and outdated systems that can slow down innovation or increase the cost of operations.
Preparation: Investing in AIOps can prove invaluable for managing technology ecosystems more effectively. This will reduce short- and long-term technology debt, while freeing up resources to drive innovation.
Regulatory landscape
TREND V. Escalating regulation: global lawmakers are working to control the power of AI.
Prediction: The AI Act introduced by the European Union could result in the first fine related to artificial intelligence. The EU is spearheading the introduction of increasingly restrictive regulations on the technology. This could discourage investment in AI within the EU and stifle growth, or alternatively become a model for legislation in the US, UK and Asia-Pacific, which the AI research communities are still waiting for with bated breath.
Preparation: Implement robust data governance and adequate security measures, prioritise transparency as well as comprehensibility of AI algorithms and establish clear accountability structures for AI-related decisions.
Please find the link to the 2025 UiPath AI and Automation Trends here.