Financial services businesses are looking to adopt generative AI to reduce costs, increase revenues, and enhance customer satisfaction, but there are significant risks involved, especially in the financial sector. TCI Media recently organized a session with EY experts to assist FinServ practitioners in starting their responsible AI journeys.
– EY experts from the Americas discussed the challenges faced by financial services firms in adopting AI, including GenAI, and suggested steps for companies to begin.
– Various regulations, such as the EU’s AI Act and around 200 data and AI bills in the US, impact GenAI adoption globally and at the state level.
– Federal regulators are learning about GenAI, with agencies like the US Treasury and CFPB seeking feedback from the sector.
– Aligning with the NIST is vital, although it may not encompass all financial services regulations, such as model risk management and cybersecurity.
– Responsible AI development involves multiple aspects like model risk management, regulatory compliance, data security, and business continuity.
– Ethical AI principles are crucial, but conflicts exist among privacy, fairness, explainability, accuracy, and security requirements.
– Practitioners face challenges in predicting the behavior of GenAI systems in operational environments.
– Building responsible AI systems requires considering metrics, toll gates, and user guardrails throughout the development lifecycle.
– Prompt engineering and having a human in the loop are essential for ensuring the responsible behavior of GenAI systems.
私の考え:
金融サービス業界において、AIの導入に関連するリスクや法規制への適合が重要であり、EYの専門家によるアドバイスは価値があると感じます。GenAIの採用は急速に進展しており、責任あるAIの開発は複雑なプロセスであり、倫理的な側面も重要です。将来的な進展に向けて、これらの課題に焦点を当てることが重要であると考えられます。