Responsible AI and Government: High Time to Open Discussions?

While businesses may not understand the need for adopting Responsible AI, can government help?

Recently, Gartner released a series of Predicts 2021 research reports, including one that highlights the serious, wide-reaching ethical and social problems it predicts artificial intelligence (AI) to cause in the next several years. The race to digital transformation and abundance of data has coerced companies to invest in artificial intelligence technologies. Even the COVID-19 pandemic acts as perfect catalyst for this! And with that, the concept of leveraging responsible AI took central stage in discussions between government, enterprises and other tech purists and critics.

A quick search trends shows that the words like "Ethical AI", and "Responsible AI" have gained popularity in the past five years. But what is the reason behind it? Currently, presence of bias in training data for artificial intelligence models and lack of transparency (black box) threaten the possibility of using AI for good. While explainable AI and trustworthy AI promise to fill in the gaps, they alone are not sufficient.

We have come a long way since 1955 when John McCarthy, winner of the Turing Prize in 1971 defined artificial intelligence as "Making a machine behaves in ways that would be called intelligent if human were so behaving". So, if we intend to use this disruptive technology for social good and fairly bring change in industries and society, it is time to switch to Responsible AI. This form of AI aims to consider the ethical, moral, legal, cultural, and socio-economic consequences during the development and deployment of artificial intelligence systems. In short it accommodates and encourages pursing AI from ethical, accountable, interpretable, qualitative, and transparent ground – something that beneficial AI and ethical AI lacked!

Why do we Need Responsible AI?

Responsible AI can refer to vast trove of factors like eliminating model bias, enhancing data privacy, fair pay for members of the AI supply chain, and more. While it is more than defining parameters of technological discipline, many have not understood the why misuse of artificial intelligence can be irreversible and problematic. As per a 2018 global executive survey on Responsible AI by Accenture, in association with SAS, Intel and Forbes, 45% of executives agree that not enough is understood about the unintended consequences of AI.

Therefore, with sharp increase in use of artificial intelligence in business, the clarion call for Responsible AI has surged in academia, industrial and government circles. While there are certain trade-offs that need to be considered, government bodies can play a crucial role to push Responsible AI into business framework. With 2021 touted as the year when Responsible AI will be an important tech trend, a strong Responsible AI framework focuses on mitigating AI risks with imperatives that address four key areas. These are: Governance, Design and Implementation, Operation and Monitoring, and Reskilling. Further, apart from the administrative and legal process, an ethical code must be devised that emphasizes on data protection and IP, privacy, transparency in decision making and mitigating social dislocation.

At present, Canada continues to be a leader and shares their Responsible use of AI in their government operations. Last year, in December, AI Global in collaboration with the World Economic Forum and non-profit Schwartz Reisman Institute, had convened the first meeting on a new program for Responsible AI Certification (RAIC). Even Prime Minister of India, Narendra Modi stressed on the fact that Algorithm Transparency is key to establishing trust in how artificial intelligence is used and it remains our collective responsibility to ensure it. Addressing viewers at virtual Responsible AI for Social Empowerment (RAISE) Summit 2020, he also announced that the government has launched Responsible AI for youth programmes this year in April.