AI trust needs more than accuracy, better evaluation: IIT Madras AI dept head

IIT Madras professor Balaraman Ravindran questioned accuracy-only assessments of AI at ConfAI 2026. In his words, he highlighted the requirement for fairness and better evaluation that could ensure trust in AI.New Delhi,Nov 5, 2026 13:48 ISTArtificial intelligence systems cannot be judged by their accuracy alone, as their responses may change depending on how questions are framed, Professor Balaraman Ravindran, Head of the Department of Data Science and AI at IIT Madras, said at ConfAI 2026, held at Plaksha University.In his keynote address on AI governance, Ravindran examined the limitations of existing methods used to assess AI systems. He highlighted why accuracy alone cannot determine whether an AI system is trustworthy, particularly when its responses may vary with the framing of questions.His presentation drew attention to the need for more rigorous evaluation of safety, fairness and reliability as AI begins to influence decisions across different fields.RESEARCHERS DISCUSS CHALLENGES IN MAKING AI RELIABLEThe three-day conference, which concluded on November 1, brought together more than 500 researchers and industry professionals inclusing Google's DeepMind Vice-President Dr Manish Gupta, who discussed advances in foundation models and their possible applications, including AI systems that understand different languages and cultures and the use of satellite imagery in agriculture. He also spoke about the challenges of making these technologies safer, fairer and more accessible.During an interaction with students, Gupta emphasised the importance of "learning to learn, always", asking them to remain curious and adapt as technology changes. Professor Rudra Pratap, Vice-Chancellor of Plaksha University, said, "India needs to build its own AI capabilities, not merely adopt technologies developed elsewhere. Universities must connect research across disciplines with the real problems our society faces."The conference also featured Professor Parag Singla of IIT Delhi, Professor Ponnurangam Kumaraguru of IIIT Hyderabad, Padma Shri awardee Dr Sanghamitra Bandyopadhyay, former Director of ISI Kolkata, and Professor Santanu Chaudhury of Ashoka University.Research presentations covered healthcare, robotics, urban systems and responsible AI. These included work on the early detection of Alzheimer's disease, robots that learn from human demonstrations, and the use of AI in drug design and urban planning.A panel titled "Beyond Chatbots: Bridging the Gap Between AI Use and Real-World Impact in India" examined the difficulties of deploying AI in healthcare, agriculture, education and small businesses.Participants discussed how limited data, differences in languages and resource constraints can affect the use of AI systems outside controlled environments.- Ends