As momentum grows for AI governance, what can companies do?

October 2023

---

AI is rapidly advancing and becoming an integral part of our daily lives, from recommendation systems on social media to autonomous vehicles and healthcare diagnostics. As AI’s capabilities grow, so do the ethical, legal, and societal challenges it presents. Effective governance frameworks for AI are therefore crucial to ensure that it is developed, deployed, and used in a responsible and accountable manner. 

Although there is broad agreement about the urgency of creating a governance framework, there is little alignment on what should be done, how to do it, who should do it, and – most importantly – whether it should be done. 

In fact, some of the companies at the forefront of AI, such as OpenAI and DeepMind, have been highlighting the risks around AI and in March of this year, they and a number of other concerned groups went as far as to call for a moratorium on the training of large AI models. 

Meanwhile, European Commission President Ursula von der Leyen became so concerned that she devoted part of her recent State of the Union address to the risk of extinction from AI. 

Von der Leyen offered up the Intergovernmental Panel on Climate Change (IPCC) as a potential model for a future, expert-led panel on AI, while others have touted the Civil Aviation Authority as a possible role model. 

The need for AI governance frameworks 

The need for AI governance frameworks can be broadly broken down into a number of key areas: 

Ethical concerns: AI systems can make decisions that have significant ethical implications, raising questions about fairness and bias. 

Bias and fairness: Many AI systems are trained on historical data that can perpetuate biases. Without proper governance, AI can reinforce and exacerbate societal and cultural biases, leading to unfair outcomes. 

Privacy issues: AI often requires access to large amounts of personal data. Without proper governance, this can pose privacy risks, including data breaches, identity theft, and surveillance. 

Accountability and transparency: When AI systems make decisions, it may be unclear who is responsible if something goes wrong. Governance frameworks are needed to define accountability and liability. Transparent AI is also essential to ensure that individuals and organisations can understand and trust AI-driven decisions. 

Ways of providing governance frameworks 

To address these concerns, various approaches have been put forward to provide governance frameworks for AI. 

The EU’s High-Level Expert Group on AI has proposed seven key principles: human agency and oversight, technical robustness and safety, privacy, and data governance, transparency, diversity, non-discrimination, and accountability. 

National and regional governments are actively working to establish legislation and regulatory frameworks for AI in order to provide legal guidelines and requirements for the development and deployment of AI systems. Again, the European Union has been developing the Artificial Intelligence Act, which aims to regulate AI applications in various sectors. 

Meanwhile, in the United States proposals for AI regulations are emerging in areas like autonomous vehicles and facial recognition technology. 

Certification and Standards 

Many see the development of AI certification and standards essential to promote responsible AI practices. International organisations, industry associations, and governmental bodies are currently working to create standards that guide AI development and use. 

Effective data governance is also essential to responsible AI development. AI relies on vast amounts of data, and its handling, processing, and storage must comply with legal and ethical standards. 

The role of companies in AI governance 

Amid all the macro-level work going on, companies also have a significant role in shaping the governance of AI. 

Companies will need to establish and follow ethical AI practices that prioritise fairness, transparency, accountability and data privacy. This includes identifying and mitigating biases, ensuring data protection, and providing mechanisms for auditing AI systems. 

AI impact assessments will be needed to evaluate the social, economic, and ethical implications of AI technologies. These assessments will help in identifying and mitigating potential risks and promoting responsible AI development. 

Companies can also advocate for ethical AI by participating in industry associations, working groups, and policy discussions and prioritise transparency by providing clear documentation on how AI systems make decisions. 

The importance of global governance 

AI is a global technology, and issues related to AI governance transcend national boundaries. For effective AI governance, international collaboration is crucial. Governments, industry associations, and international organisations will need to work together to create a cohesive global framework and establish common ethical standards and best practices. 

Governance will be at the heart of how AI is adapted into all walks of business life, yet while a clear path towards common practices and standards has been flagged as a matter of urgency, bringing all industries, geographies and policy makers to a common cause could be a challenge tough enough to confound even the most intelligent robotic mind.