The panel discussion on fueling technological innovations through efficient regulations of big data was organised by the Berkeley Law Society (BLS) on April 17, 2019 at The A, One Indiabulls Centre, Mumbai.
The discussion was a multi-stakeholder initiative of bringing technologists and lawyers together to bridge the gap between emerging technologies and policies. The panel comprised of representatives of data-intensive and AI centric organisations: Varun Puri, CPO + Co-founder of inFeedo; Vinay Kumar, Founder & MD of Arya.ai; Shreya Kolay, Partner of Bankura Seva Niketan Hospital and Karthik Kannan, Co-founder of Youcode Intelligence Solutions. The panel was moderated by Sharmila Nair, a BLS member and lawyer specializing in Technology and Intellectual Property Laws.
The discussion focused on –
Internal policies, if any, used by stakeholder groups to regulate the big data space for better performance of their product/service;
To identify the common thread as well as functional differences faced by stakeholder groups to increase accountability and security for big data collected and analysed; and
Methods of democratising data for greater cross-sectoral efficiency.
The aim was to understand the regulatory concepts of ‘security’, ‘transparency’ and ‘accountability’ from the technologists’ perspective so that the concepts eventually gets translated into the motto of #AiforAll.
Varun addressed issues of internal privacy guidelines if Amber – it’s closed domain chatbot was converted into an open domain bot requiring more stringent guidelines. In the course of discussions, Vinay confirmed that Arya was incoporating explainability into the design stage of deep learning systems. This has often been a worry with ‘black box’ algorithms.
Ensuring accountability with emerging technologies is a matter of grave concern as emphasised by Shreya, especially for the medical sector. Karthik was concerned that without national guidelines/laws, regulating emerging technologies would be difficult.
The panel also addressed growing security concerns which could be regulated by design and policy due to the fluidity and secondary use principle of big data. Data silos can impede the cross-sectoral impact of data. The impact of risk assessments and extent of human involvement, while dealing with sensitive data was analysed in light of the Algorithm Accountability Bill and EU’s AI Ethics Guidelines.
The audience showed their concern through their interactions and felt that the panel was a step in the right direction of regulating technologies efficiently especially with multiple stakeholders involved.