The EU made a significant move towards regulating artificial intelligence (AI) by tentatively agreeing on the terms of the AI Act.
This legislation, subject to approval by member countries and parliament, is poised to set global standards by adopting a risk-based approach to AI regulation. RegTech firm Saifr recently outlined how proposed AI regulations take a risk-based approach to help ensure innovation, safety, and a healthy financial ecosystem.
According to the firm, such an approach is particularly pragmatic, acknowledging the varied applications of AI across industries, including the financial services sector. Under the proposed framework, AI systems are categorized based on the level of risk they pose, ranging from minimal to unacceptable.
AI applications with minimal risk, like email sorting, are exempt from regulatory oversight. Conversely, high-risk AI systems, particularly those affecting health and safety, will be subject to stringent requirements, including robust risk-mitigation systems, high-quality data sets, and human oversight. The act also addresses AI systems posing unacceptable risks to human rights, categorically banning them, and mandates full transparency, ensuring users are aware when interacting with AI.
In the view of Saifr, financial services, an industry well-acquainted with regulations from bodies like the SEC and FINRA, understands the importance of regulatory oversight in maintaining market integrity and trust. Imagine a market devoid of regulations. Investors would seek assurance of truthfulness, transparency in risks and returns, fairness in transactions, and security of assets. Regulations provide this assurance, fostering an environment where participants can engage confidently and securely.
Similarly, as AI technology integrates into financial services, new regulations are imperative to guide its responsible adoption and mitigate potential negative impacts. Saifr believes that in essence, just like financial regulations, AI regulations are set to become the guardrails that ensure the industry’s healthy progression.
AI, akin to any emerging technology, brings a blend of potential risks and rewards to the financial services sector. Historical examples, like the advent of cryptocurrency and blockchain, illustrate the volatility that can arise from regulatory uncertainty. However, they also demonstrate the transformative potential of new technologies in enhancing transparency, efficiency, and user control.
Proper regulations are crucial, not just for the safe adoption of these technologies but also for their integration into regulated industries like financial services. Learning from the past, it’s clear that regulated AI can be safely integrated into financial services, offering benefits such as improved compliance, fraud detection, and enhanced customer engagement.
Firms like Saifr are at the forefront of leveraging AI to reinforce regulatory compliance within financial services. Saifr introduces an additional layer of AI, acting as guardrails to ensure that outputs from AI systems, such as those used in marketing or customer interaction, align with regulatory standards.
Large language models excel at tasks like analysing customer data and generating personalized content. However, without an understanding of the regulatory landscape, the content produced might not meet the required compliance standards. Saifr’s AI scrutinizes the content, flags potential compliance risks, and suggests adjustments, aiding humans in maintaining regulatory compliance across text and visual content.
Read the full post here.
To read the latest breaking RegTech stories as they happen visit https://member.regtechanalyst.com/
Copyright © 2024 RegTech Analyst
Copyright © 2018 RegTech Analyst
