Equity Group Develops AI to Forecast Staff Departures

Equity Group, Kenya’s second‑largest bank by assets, says it is developing an artificial intelligence tool to forecast employee departures as part of efforts to retain specialist staff and give managers more time to fill critical roles. The work is described in the group’s 2025 sustainability report as predictive analytics intended to identify risk windows and allow earlier intervention.

What Equity is doing

The report says the lender sees competition from fintechs, telecommunications companies and rival banks for digital and payments talent, making retention as important as recruitment. Equity nearly doubled hiring to 3,198 people in 2025, raising total headcount by 2.2% to 13,370. Its permanent workforce increased by nearly 1,000 while temporary staff fell by almost 700. Net employment turnover fell to 10.6% in the year to 2025 from 11.1% in 2024; the previous year saw a larger drop from 17% to about 11%.

Staff costs rose 18% to KES 39.5 billion ($303.9 million) from KES 33.4 billion ($257 million). The group also reports that regional banking subsidiaries accounted for 48% of banking profit after tax in 2025, with the Democratic Republic of Congo contributing KES 24.7 billion. To support skills and mobility, Equity already seconds specialists from its Nairobi head office to subsidiaries, is consolidating workforce records and performance data into a single human resources system, advertises vacancies internally first, and is formalising career paths that let technical specialists progress without moving into management. Average training time rose to 74 hours per employee in 2025 from 45 the year before, and some generative AI courses have been made mandatory for staff.

Implications for developing countries

The move illustrates how financial groups operating across African markets are turning to analytics to manage scarce technical talent and reduce disruption to regional operations. Predictive tools could help target retention and training investment where skills are concentrated, but the report gives no launch date and leaves unanswered how the model will operate, what employee data it will use, and whether it will identify individuals or forecast at team or market level — gaps that have implications for worker privacy and fairness.