Rotterdam school of Management, Erasmus University compact logo

Article: Thursday, 24 March 2025

This blogpost from Sheila Gemin, Global Head for Corporate Technology at ING and Lesly Nzeusseu, a doctoral student in Industrial & Organizational Psychology at the University of Montreal, and Reshaping Work Fellow describes ING’s use of AI and considers its use in the banking sector, taken from a round table discussion at the Reshaping Work Conference in October 2024.

The financial sector has experienced a surge of interest in AI technologies, with potential applications ranging from automation of customer service to risk management. But the rise of AI brings data security and regulatory challenges into focus. At ING, a leading European bank with more than 60,000 employees and 38 million customers, these considerations shape every AI initiative. In recent multi-stakeholder discussions held at the Reshaping Work Conference 2024, participants explored how ING is leveraging AI while prioritising data security and regulatory compliance.

Historical perspective

Data analysis has long been a pillar of ING’s operations, particularly in understanding customer behaviour., yet past efforts were limited by underdeveloped legislation and an evolving understanding of data privacy. Today, as data collection capabilities have expanded, so has ING’s commitment to rigorous data privacy and compliance standards, including a cautious approach to AI tools like ChatGPT, which is currently closely monitored.

Current AI initiatives at ING

ING recognises the potential of Generative AI as a way to provide a better customer experience. ING’s aim is to interact with their customers in a personal, fast, relevant, and easy way and it is exploring innovative technology like GenAI while taking a prudent and responsible approach to doing this in a safe and secure way:

1. Copilot and chatbot development

ING has rolled out versions of GenAI for internal use on GitHub – a platform for hosting code that allows for version control and collaboration – and on Microsoft Office 365 Copilot, under strict security protocols. These AI tools aim to streamline email and document management while ensuring no vulnerable data can be exposed externally. Copilot is also used as a productivity tool.

2. Sustainability reporting and assessment

Sustainability is central to ING’s values, and the bank is exploring AI-assisted assessments of loan portfolios for environmental, social, and governance (ESG) compliance. Using internet data and advanced analytics, these assessments contribute to sustainable decision making and reporting.

3. Exploring local LLMs for enhanced security

ING is researching, investigating and cautiously deploying large language models (LLMs), while acknowledging the complexities and resources required for accurate data training. For sensitive data – such as that which includes personal identifiers at the C4-level of software architecture – ING applies extensive verification and anonymisation processes. Data involved in any external transfer is encrypted, ensuring maximum protection for client information.

Case study examples

During the round table discussions at the Reshaping Work Conference, ING representatives shared the AI applications under development:

  • Customer-focused chatbots: Generative AI-driven chatbots are rigorously tested and trained to ensure accurate, secure interactions with customers.
  • Investment advice: ING explores how AI can assist in providing insights for long-term investments, though challenges remain due to regulatory restrictions and the need for human oversight.

AI and data security

Participants from Zurich Insurance and Decathlon at the conference echoed similar experiences with AI, highlighting a universal need to balance innovation with data protection. While ING's data governance policies may limit certain AI functionalities, the priority remains clear: safeguarding client trust and data integrity. ING’s internal AI committee, which includes its CEO, carefully reviews each AI use case for potential benefits and risks, demonstrating the bank’s commitment to responsible AI integration. Lastly, one of the most notable insights from the discussion involved the cost implications of AI deployment, particularly for tools like Copilot. Due to the high costs involved, ING is cautious about expanding the rollout without clear ROI indicators. The ongoing debate surrounding productivity measurement and cost efficiency reflects ING’s commitment to sustainable, value-driven AI implementation.

Conclusion: the road ahead

As the discussion concluded, the importance of long-term strategic thinking in AI adoption was emphasised. ING’s approach serves as a case study for other financial institutions: rather than rushing to implement every new AI tool, ING takes a deliberate, security-focused approach that prioritises data integrity and compliance. With the rapid evolution of AI technology, ING’s careful balancing act between innovation and regulation could set the standard for responsible AI use in the banking sector.

RSM Discovery

Want to elevate your business to the next level using the latest research? RSM Discovery is your online research platform. Read the latest insights from the best researchers in the field of business. You can also subscribe to the newsletter to receive a bimonthly highlight with the most popular articles.
Your contact for more information:
Danielle Baan

Science Communication and Media Officer

Erasmus University campus in autumn, showcasing its iconic red trees, viewed from across the campus pool.