Convergence & Broadcasting division has released a Draft Standard for assessing and rating Artificial Intelligence Systems for fairness. Artificial intelligence (AI) and Machine Learning (ML) applications are increasingly used in all domains. Unintended biases in their predictions or outcomes are a big concern.  

The collaboration explored the dynamics relating to fairness in the Indian context, including the diversity, the protected and proxy variables, and approaching fairness in a risk-based method to ensure that the fairness assessment is tailored to the relevant use case. Furthermore, it attempts to cover multiple contributors of bias with the process, metrics and scenario (for instance, contributed by adversarial / causality) based approach.  

One important requirement of Responsible AI is that the AI/ ML System should be unbiased or fair. Therefore, in achieving the objective of the Government of India of building public trust in AI/ ML Systems (#AIforAll), TEC has initiated a Fairness Assessment of AI/ ML Systems on a Voluntary basis. 

As a consequence of the public consultations, TEC constituted a Working Group of members from Industry, Academia, Researchers, subject experts from Government departments, etc., to prepare the initial draft of the proposed Standard for assessing the fairness of AI Systems.  

C&B Division TEC held public consultation through an interactive webinar on 22nd March 2022, followed by Consultation Meeting on 1st September 2022 for framing procedures for assessing fairness for different types of AI/ML Systems, and for issuing fairness rating/ certifications to AI/ML systems, as a benchmark of fairness.  

The team is expecting input from the public. They can be shared with Avinash Agarwal in the email ID provided in the document.