Background
PPS is an 85-year-old South Africa-based mutual that focuses exclusively on graduate professionals. Membership requires a four-year degree, and the organisation serves around 140,000 members across professions such as medicine, engineering, and accountancy. Products offered include life insurance, sickness and disability cover, short-term insurance, investments, and medical aid administration.
An “iceberg” model
Tools like ChatGPT changed expectations across industries and roles. PPS initially tried to restrict AI use due to data exposure concerns, but it then recognised that employees would experiment anyway if it made work easier.
Some employees are anxious about AI potentially changing or replacing their jobs. Using an “iceberg” model lens, PPS distinguishes visible, formal adoption from the less visible behaviours and concerns that leadership and HR may not immediately detect.
AI uses
PPS has approved a primary AI tool, Microsoft Copilot, while its IT team continues to assess the security posture of other platforms. AI features are increasingly embedded in everyday applications (such as Word, Excel, and meeting tools), making it harder to separate “AI use” from normal digital work.
In HR, PPS sees substantial uptake across the employee lifecycle, particularly in attraction and recruitment, where candidates may use AI to draft CVs and applications. It also uses screening and agent-based shortlisting to speed up hiring, and it has also built an internal policy “FAQ” bot to answer staff questions on topics like leave, travel, and acceptable use.
Responsible use
A recurring concern is responsible use: AI can support work, but accountability for decisions and outputs stays with the employee, reinforcing a “human in the loop” principle. To build capability, PPS segments staff into user groups and promotes AI literacy through regular learning formats such as Lunch & Learn sessions, departmental sharing, newsletters, and town halls.
Governance implications
On governance, PPS compares common organisational approaches—from open use, to blocking, to sanctioned tools, sandboxing, and fully governed enablement—balancing risk, productivity, and implementation effort. PPS has established an AI Governance Council under an executive mandate, maintains an AI inventory and risk-tiering, and routes new initiatives through assessment and approval before publishing them in an internal “catalogue”. It has also benchmarked its approach against insights from a research and advisory firm and is strengthening monitoring, ownership clarity, reassessment practices, and vendor data-processing agreements to protect company data.
