BENGALURU: The injection and integration of artificial intelligence (AI) into digital spheres of interaction between institutions and the individual have been rapid and steep, allowing little time for the law and policy to catch up, say Bengaluru-based experts.
According to them, as the finance sector is increasingly moving towards AI-based automation without legal frameworks of accountability already in place, consumers are left to their own devices, leaving them vulnerable to making mistakes involving personal and confidential information.
When an agency offering financial advice uses AI in its operations, loopholes in the accountability of personal information are inherent in the interactions between an individual and the AI platform. In such a situation, placing the onus on the individual to enact discretion makes the situation further complicated.
“Consumers do not need to fear every AI-driven financial platform, but they should approach them with informed scepticism. AI can assist with financial information, yet decisions involving savings, debt, insurance, retirement or investments still require transparency, human accountability and independent judgment,” said Prof Madhu Veeraraghavan, pro vice-chancellor for management, law, humanities and social sciences at the Manipal Academy of Higher Education (MAHE).
What introduces further complication is that owing to the aforementioned lack of any ethical standard of procedure or similar framework, it is not uncommon for such platforms to run their customers’ queries through layers of AI, without expressly asking for consent.
“If customers are not clear on how much algorithmic systems impact their decisions due to the lack of transparency, they could watch out for some signs: highly standardised replies with minimal context, minimal human interaction, instant automated responses, etc.,” said Remya Tressa Jacob, an assistant professor in the Accounting, Economics and Finance Area department at TA Pai Management Institute (TAPMI).
The lack of a policy or governance with regards to AI also shows it is by virtue elusive to such frameworks, which should be a cause for significant concern. Prof Tuhin S Banerjee, associate dean of the School for Continuing Education and Professional Studies, RV University, said, “The inherent lack of explainability in deep learning makes it difficult for financial institutions or regulators to understand the reason for the decision, thus complicating compliance. Data misuse and unauthorised profiling may happen in trying to use AI for hyper-personalisation.”