Comptroller and Auditor General of India (CAG) Girish Chandra Murmu sees scope for devising a common international framework for audit of Artificial Intelligence (AI), which has captured the imagination of the world in numerous ways, revolutionising various industries and transforming the daily lives.
Putting in place a common global audit standard after the Supreme Audit Institutions (SAI) of G20 countries collaborate and build consensus on a set of common parameters can help the member countries approach the use of AI technology in a coherent manner, Murmu said at the end of SAI20 Summit held at Goa from June 12-14.
Murmu is currently the Chair of the Supreme Audit Institutions-20 (SAI20) Engagement Group, which organised the SAI20 Summit.
Auditing standard for AI
Asked if time has come for SAI20 countries to come together to frame a global auditing standard of AI, Murmu said that at the apex level there are already international auditing standards on various other issues.
“We already have common framework in the form of global auditing standards on various areas. But even on the area of AI also, there can be scope for a common global framework”, Murmu said.
Noting that SAI20 is a major forum representing 85 per cent of the Global GDP, Murmu said this platform was currently collaborating and doing experience sharing on AI.
“Once we come out with some kind of collaboration or experience sharing, certain common parameters will come out. Once the common parameters are out, we can then perhaps come up with a global guide on that (audit of AI)”, he added.
Policymakers everywhere from Washington to Brussels to Beijing are now grappling with the advancements of AI and racing to control this transformative technology.
European Union this week took a major step towards regulating AI. The European Parliament passed a draft law by the name A.I. Act, which seeks to introduce new restrictions on AI’s riskiest uses. This was the first step of law making with the final version expected by the later part of this year.
Meanwhile, Murmu made it clear that CAG is concerned only about audit and does not deal in policy side or regulatory side of AI. This was something that nations have to decide, he noted.
“It is for the countries (their governments) to look at the downsides and threats of AIs and build guardrails. Important aspects like data security, personal security and authenticity part needs regulation so that AI does not generate misleading information”, he said.
Blue economy
Murmu noted some regulatory changes may be required in India on AI, but none whatsoever for the Blue economy. He however noted that some standardised audit guidelines are required for blue economy.
Asked if the SAI20 was particularly focused on the advancements in generative A.I, Murmu replied in the negative. He said that SAI20 was focused on bigger theme of Responsible AI.
At the recently concluded SAI20 Summit, the Supreme Audit Institutions of G20 countries agreed to cooperate and collaborate in knowledge sharing endeavours involving exchange of innovative ideas and best practices, thus equipping SAIs to face the challenges of auditing the frontier spheres of ‘Responsible Artificial Intelligence’ and ‘Blue Economy’.
In a compendium on AI released during the SAI20 summit, Murmu as the chair of the knowledge Sharing Committee of INTOSAI and the working group on IT Audit recommended working towards setting guidance for adopting AI in audit and also developing auditing framework for optimising the use of AI in audit and promoting responsible AI for good governance.
He highlighted that SAIs must adapt their skills and techniques to be able to audit AI algorithms in audits.
Today AI has reached a level where AI could contribute upto $ 15.7 trillion to the global economy by 2030, Murmu noted in the compendium.
(This correspondent was recently at Panaji at the invitation of CAG of India for SAI20 Summit).