Kiwi FSIs Find Third-Party Tech Better Than In-House AI Development

New Zealand’s financial services industry has experienced a notable shift in its approach to artificial intelligence (AI). Initially, many firms were eager to develop their own proprietary AI Development Platforms. However, recent findings from the Financial Markets Authority (FMA) reveal that this initial excitement has not fully materialized. Instead, financial service institutions (FSIs) are increasingly turning to third-party AI solutions, including various AI development platforms, rather than investing heavily in in-house developments.

The FMA’s latest report, based on a comprehensive survey of New Zealand’s financial firms, provides insight into how these organizations are adopting AI technologies. The report highlights a cautious approach to AI, with firms primarily using AI for risk management and fraud detection, rather than for broader business applications.

Of the 13 regulated entities surveyed—including banks, insurers, asset managers, and financial advisors—at least nine are actively utilizing AI technologies in their operations. However, their use of AI is largely focused on risk identification, assessment, and mitigation. The integration of AI into other business areas remains limited, underscoring a careful and measured approach.

Insurers and financial advisors are using AI to improve operational efficiency and customer service. For example, insurers are employing AI to streamline processes and enhance client interactions, while financial advisors leverage AI to offer personalized advice. In contrast, deposit-taking institutions are focusing their AI efforts on risk management, particularly in fraud detection. This focus on risk management is less evident among insurers and financial advisors.

A significant finding from the FMA report is the cautious stance towards generative AI. While there is growing interest in AI, the financial sector predominantly uses simpler machine learning tools rather than advanced deep learning systems. Many firms are still exploring how to safely implement generative AI technologies, reflecting a thoughtful approach to AI application.

Despite the preference for third-party AI solutions, the FMA report highlighted an interesting case where a firm successfully developed its own AI application development platform. This proprietary platform, akin to an Open AI developer platform, was created to protect internal data and to better fit the firm’s specific needs. The decision to develop an in-house AI solution was driven by the need to keep sensitive information secure and to customize the platform to the firm’s unique requirements.

The survey identified several current AI use cases across the financial sector:

1. Efficiency Tools: Off-the-shelf tools, such as Microsoft CoPilot and GitHub, are being tested to enhance team efficiency, support product development, and assist with documentation and knowledge management.
2. Security and Fraud Detection: Machine learning tools like Darktrace, Databricks, and HuggingFace are utilized for security purposes, including fraud detection, pattern recognition, and predictive modeling.
3. Web Automation: Tools such as Miro and Zapier are used for automating processes like credit decision-making, document management, and analyzing customer behavior.
4. Security Tools: AI-driven security solutions, such as Darktrace and Egress, use self-learning algorithms to detect and respond to cyber threats by identifying anomalies and deviations from normal behavior.
5. Customer Service: Chatbots are employed to improve customer service by summarizing interactions, offering search options, and drafting documents in a user-friendly manner.

Looking ahead, New Zealand’s financial firms are motivated to adopt AI to enhance customer outcomes and operational efficiency. While fraud detection remains a key use case, it is less emphasized as a future motivation since many firms have already established systems for it.

FMA Chief Economist Stuart Johnson highlighted that the survey aimed to uncover both the benefits and risks associated with AI. He emphasized the need for a balanced approach to AI, recognizing its potential while addressing governance and risk concerns. Johnson noted that AI offers significant opportunities but also presents emerging risks that require effective management.

The FMA remains committed to supporting technological advancements and ensuring responsible adoption. They are technology-neutral and pro-innovation, aiming to ensure that New Zealanders have access to the same technological benefits as those in other countries. To further explore how AI and generative AI are being used in New Zealand’s financial sector, the FMA will host a roundtable on October 1, 2024. This event will bring together industry participants to discuss AI applications, including the use of various AI development platforms, and strategies for managing associated risks.

In summary, while New Zealand financial services firms initially pursued the development of their own AI platforms, the trend is shifting towards adopting third-party AI solutions. The industry’s cautious approach reflects a focus on risk management and operational efficiency, with ongoing exploration of advanced AI technologies and platforms. The FMA’s report underscores the importance of balancing innovation with robust governance to effectively harness the benefits of AI.