BSI research has revealed that guidance on Artificial Intelligence (AI) produced to date by international bodies, national Governments and technology providers alike is “disconnected”. As a result, the business standards company is calling for the consolidation of standards and “regulatory consistency”.
AI presents ethical and social challenges that require a global and standardised response, according to the BSI. In its new White Paper entitled ‘Overview of the Standardisation Landscape in Artificial Intelligence’, the organisation suggests that current guidance “includes disparate and repetitive advice”.
Over 40 AI guidance papers have been published or developed over the last two years, with international bodies, including the OECD, International Standards – ISO/IEC, national initiatives (including ones in Singapore and the UAE) and global technology providers producing sometimes conflicting advice.
Many of the papers provide guidance on the same core themes including safety, ethics and governance. The BSI’s White Paper suggests that technology policymakers and businesses people alike are unable to keep abreast of the guidance and are not able to navigate it easily.
The BSI wants to see standard consolidation and regulatory consistency to avoid duplication of both the workload of producing the guidance and the output papers.
Tim McGarr, digital sector lead at the BSI, stated: “With the rapid emergence and further development plans for AI, transparent and trustworthy ways of working will be central to the technology’s success. Our White Paper shows the volume of standardisation and standards-like activities taking place across the international market. It aims to help the industry create consensus-driven standards that will help us face the challenges ahead.”
McGarr added: “We’re calling on the industry to come together to incorporate all voices into the governance of AI technologies. Now is the time for key stakeholders, including Government, regulators, industry and consumers to engage in the standards development process. This will ensure that we can build a successful future for AI and one which helps to protect citizens from the unintended consequences of the disparate guidance.”
Milan Patel, chair of the Committee which feeds the UK view into international standards from ISO/IEC, added: “The White Paper shows the scale of activities taking place across the industry to develop Best Practice in AI. It highlights the need for multiple stakeholder collaboration on key standardisation areas to aid the trustworthy development and use of AI.”
*Download the BSI’s White Paper here: https://www.bsigroup.com/en-GB/Aerospace/artificial-intelligence/download-the-artificial-intelligence-whitepaper/