ع
Slawomir Krupa told attendees at the Morgan Stanley European Financials Conference that high supervisory expectations around AI tools create a substantial burden of proof especially for applications touching clients or risk management. The chief executive noted that while banks including Societe Generale conduct numerous experiments the level of reliability combined with regulatory demands has so far limited major applications at scale in production. Krupa added that this environment will slow real adoption for a while yet the technology remains poised to deliver significant change.
The French bank established a dedicated entity called SocGen AI to industrialize use cases and approach development from an external perspective free of legacy constraints. In an announcement on its newsroom Societe Generale said the unit coordinates initiatives across six domains ranging from client interaction and onboarding to compliance and productivity gains. The chief executive described the bank as a large digital factory that processes data continuously making AI a transformative force despite the regulatory pace.
Societe Generale signed a strategic agreement with Anthropic to accelerate corporate AI adoption the bank said in a statement distributed via GlobeNewswire on September 21 2026. That announcement formed part of a broader strategic roadmap targeting gross savings of 1.9 billion euros by 2029 with AI expected to generate between 500 million and 600 million euros of those efficiencies. Reuters reported that the plan also seeks to reduce the cost-to-income ratio below 55 percent while lifting return on tangible equity to between 13 and 14 percent in 2029.
The bank’s statement detailed that AI productivity measures would help cut IT spending by around 0.5 billion euros lowering the IT intensity ratio to 12 percent by the end of the period. Additional savings would come from procurement reductions and a gradual decline in employee numbers through natural attrition. Krupa stated that the group had already simplified its model and strengthened discipline over the prior three years positioning it for the next phase of value creation.
Societe Generale maintains roughly 200 AI-based use cases in production according to its digital and innovation overview. These deployments include chatbots at BoursoBank and SG in France that manage more than six million client interactions each year on a 24/7 basis. The bank stressed that its AI governance embeds ethical principles regulatory compliance and bias management to build trust across stakeholders.
Krupa’s remarks arrive as European Union regulators implement the AI Act which subjects many banking systems to high-risk classification and strict oversight requirements. The lender’s partnership with Anthropic focuses on safety-aligned models suitable for regulated environments the statement said. Industry data from the bank’s disclosures show that responsible deployment of such tools supports both efficiency and resilience in core operations.
