Transforming Professional Indemnity insurance
As the global insurance industry embraces rapid technological change, Artificial Intelligence (AI) is becoming a transformative force - particularly in the field of Professional Indemnity (PI) insurance. From enhancing underwriting precision to reshaping the skillsets of professionals, AI presents insurers with immense opportunity - balanced, however, by complex challenges involving data integrity, legal compliance, and human oversight.
To understand how AI is reshaping this space, FAnews consulted legal and insurance experts from leading firms. Bianca Da Costa and Venolan Naidoo, Partners at Fasken, as well as Jean-Paul Rudd and Mtho Maphumulo, Partners at Adams & Adams, and Dasran Padayachee, Claims Team Manager at Santam, share their insights on what insurers must know to adapt, comply, and thrive in an AI-driven future.
AI in risk assessment
AI is fast becoming an essential tool in identifying, evaluating, and predicting professional indemnity risk. Traditional manual underwriting processes are giving way to automated, data-driven methodologies.
Fasken’s Da Costa and Naidoo explain that AI is now being used “to reduce manual labour, minimise delays and increase accuracy when dealing with claim management and processing procedures, risk evaluation and fraud detection.” Technologies such as Insurtech, Blockchain, Telematics, Internet of Things (IoT), Chatbots and Virtual Assistants are helping automate complex processes that previously required manual human intervention.
According to Rudd, Maphumulo and Padayachee, “AI is being used to automate and enhance the assessment of risk through analysing big data and the identification of claims trends to improve underwriting accuracy. Tools such as Generative AI and Large Language Models are leading the transformation.”
These tools enable underwriters to quickly identify anomalous patterns in professional behaviour and claims activity - leading to better-informed underwriting decisions.
The benefits of AI
The most immediate and tangible benefit of AI integration is operational efficiency. AI systems are enabling underwriters to process applications faster and with greater consistency.
Da Costa and Naidoo note that “AI tools certainly streamline the underwriting process - more accurate underwriting output, mitigate against human biases, enhance the underwriting process, lower fraud risk and reduce administrative costs, which could lead to lower premiums for customers.”
Rudd, Maphumulo and Padayachee echo this sentiment, adding that “AI enables insurers to respond more quickly to emerging risks - enhancing customer experience by enabling faster turnaround times, more personalised coverage recommendations, and reduced friction in the application process.” Additionally, the automation of repetitive tasks yields “significant cost savings” and a more agile underwriting environment.
At a time when clients expect speed, transparency, and personalised service, these technological advancements are making it possible to meet those demands without sacrificing quality or compliance - assuming the underlying data is reliable.
Legal and ethical pitfalls
Despite the benefits, experts caution against blind trust in AI systems - especially in high-stakes sectors like PI insurance, where claims can be complex and reputational damage severe. Recent court decisions in South Africa have highlighted the risks associated with over-reliance on generative AI tools without adequate human oversight.
Rudd, Maphumulo and Padayachee reference the case of Mavundla v MEC (2025), where the Court issued a punitive costs order against an attorney who relied on ChatGPT to draft legal submissions that cited non-existent cases. The judgment noted: “AI tools can produce plausible but entirely false legal citations, which may mislead courts and undermine the integrity of legal proceedings.” The Court further ruled that even negligent use of AI in legal work may constitute serious professional misconduct.
Another landmark case, Northbound Processing (Pty) Ltd v South African Diamond and Precious Metals Regulator (2025), reinforced these dangers. “The heads of argument submitted contained multiple fictitious case citations - later confirmed to be AI-generated ‘hallucinations.”
For insurers, these cases are more than legal anecdotes - they’re cautionary tales. As Rudd, Maphumulo and Padayachee note, these judgments raise “key challenges including the risk of AI-generated misinformation leading to negligent advice or flawed underwriting decisions, exposure to liability and regulatory obligations to ensure AI use complies with data protection laws and professional standards.”
Similarly, Da Costa and Naidoo warn that “insurers must ensure that when integrating AI tools, they comply with the legal requirements associated with the use and protection of such data.” With South African laws like POPIA (Protection of Personal Information Act of 2013) and section 71, which regulates automated decision-making, compliance is not optional - it’s imperative.
The human element
While AI is automating many traditionally manual processes, it is not eliminating the need for human professionals. Instead, it's redefining their roles - pushing them toward strategic, oversight, and governance functions.
According to Da Costa and Naidoo, “advancements of AI have impacted more on menial/repetitive tasks, but enhanced professionals using AI tools to be more effective in their roles.” The emphasis now lies in human oversight, understanding AI limitations, and using AI output as a support tool rather than a sole decision-maker.
Looking ahead, Rudd, Maphumulo and Padayachee foresee AI playing a greater role in policy renewals, claims assessments, and customer-specific policy design. They predict the use of “smart contracts that will result in automatic policies and faster cover for customers.”
But with that, increased automation comes greater responsibility. As Da Costa and Naidoo put it, “humans are still required to be accountable for the AI tools and the future technological changes in the industry.”
To that end, both sets of experts advise insurers to prepare for stricter regulation. Da Costa and Naidoo suggest insurers adopt “partial self-governing policy measures” aligned with international best practice, while awaiting industry-specific regulation from the Financial Sector Conduct Authority (FSCA). Until then, “insurers would undoubtedly need to navigate the current legislative landscape while using AI tools.”
A future built on balance and accountability
The integration of AI into professional indemnity insurance is no longer a futuristic concept; it is a present-day reality with tangible benefits and risks. As the technology matures, insurers will need to maintain a delicate balance: leveraging the speed and scale of AI while ensuring fairness, transparency, and compliance.
AI-driven transformation is inevitable, but as we've seen in legal rulings and expert warnings, it cannot be left to run on autopilot. The most successful insurers will be those who not only invest in cutting-edge tools but also build robust governance frameworks, promote human oversight, and stay ahead of an evolving regulatory environment.
As Rudd, Maphumulo and Padayachee aptly put it, “These judgments serve as a cautionary tale and a call to action for insurers to develop robust governance frameworks around AI use.” The call is clear - those who lead responsibly will define the next era of PI insurance.
Writer’s Thoughts
The shift toward AI-enabled underwriting demands not only smarter systems, but smarter stewardship. For insurers, the path forward lies in leveraging AI’s advantages while ensuring that accountability, compliance, and ethical practice remain firmly at the centre of every decision. Please comment below, interact with us on Twitter at @fanews_online or email me your thoughts [email protected]
