OECD Finds Countries Making Progress Implementing AI Principles, But Challenges Remain

Editor’s Note: This news story summarizes a recent report from the Organisation for Economic Co-operation and Development (OECD) titled “The State of Implementation of the OECD AI Principles Four Years On.” The OECD is an intergovernmental organization with 38 member countries founded in 1961 to stimulate economic progress and world trade. In 2019, the OECD established principles to guide the trustworthy and responsible development of artificial intelligence (AI) systems. This new report assesses global progress in putting those principles into practice. It offers valuable insights for cybersecurity, information governance, and eDiscovery professionals on emerging trends in AI governance, risks, and regulatory approaches.

Industry Article

OECD Finds Countries Making Progress Implementing AI Principles, But Challenges Remain

ComplexDiscovery Staff

Four years after the Organisation for Economic Co-operation and Development (OECD) established landmark principles to guide the responsible development of artificial intelligence (AI), a new OECD report finds countries are making significant strides to implement them but still face challenges fully realizing the principles in practice. This overview of that new report titled “The State of Implementation of the OECD AI Principles Four Years On”  summarizes the OECD assessment of the state of implementation of its AI principles globally based on a survey of initiatives in member countries as of May 2023. It covers the key trends identified around national AI strategies, emerging regulatory frameworks, and policies exemplifying the OECD’s values-based principles for trustworthy AI.

Growth of National AI Strategies Documented

A major trend documented is the proliferation of comprehensive national AI strategies to steer the technology’s growth. In 2017, only Canada, Finland, and Japan had such plans. Today, the OECD has cataloged over 50, spanning countries in the Americas, Europe, the Middle East, Africa, and Asia-Pacific.

These national strategies focus on priorities like expanding AI R&D, building digital infrastructure to support AI systems, assessing AI’s impacts on jobs and society, and monitoring the effectiveness of AI policies.

Emerging Regulatory Frameworks Highlighted

The OECD also documents emerging AI regulatory approaches to mitigate risks and biases. Many countries have now drafted national AI ethics frameworks that encapsulate principles like accountability and fairness. Some countries like Japan and Korea provide specific guidelines for organizations to implement them.

Pioneering governments are beginning to translate principles into binding laws and rules tailored to AI. For instance, Canada and the European Union have proposed comprehensive AI regulations mandating transparency and risk management practices for sensitive technologies. China and the United States are opting for more sectoral oversight.

OECD AI Principles Put Into Practice

The report includes numerous examples of policies that exemplify the five OECD values-based AI principles in action. These policies include: 

• Accountability: Requiring EU AI providers to maintain up-to-date documentation proving their systems function correctly.
• Transparency: Amsterdam and Helsinki releasing public registries of algorithms used by their city governments.
• Fairness: US and UK initiatives to understand and mitigate biases in algorithmic systems.
• Robustness: Austria legally regulating self-driving vehicles on public roads.
• Well-being: Germany funding projects applying AI to environmental challenges like optimizing rail transport and reducing greenhouse gases.

Progress Made, But Work Remains

The OECD survey shows member countries have made meaningful strides in aligning policies with the organization’s AI principles. National strategies, emerging regulations, and concrete initiatives exemplify this progress.

However, fully implementing the principles in practice remains a work in progress. More efforts are needed to manage rising risks from increasingly powerful AI systems and collaborate across borders on AI governance.

For cybersecurity, information governance, and eDiscovery fields, this report highlights valuable insights on unfolding trends in AI risks, regulations, and governance. Understanding these developments can help inform strategies in those areas that account for AI’s rapid evolution.

As countries progress in developing AI governance, the OECD principles can provide an ethical foundation for them. The principles aim to steer AI in a direction that benefits societies while protecting rights and values. However, achieving the full vision of trustworthy AI outlined in the principles will still necessitate ongoing, collective commitment from nations across the globe.

Article Sources

• How countries are implementing the OECD Principles for Trustworthy AI – OECD.AI
• OECD (2023), “The state of implementation of the OECD AI Principles four years on”, OECD Artificial Intelligence Papers, No. 3, OECD Publishing, Paris, https://doi.org/10.1787/835641c9-en.

Assisted by GAI and LLM Technologies

Additional Reading

• The Green Equation: How ESG and Green Computing May Boost the eDiscovery Bottom Line
• Weighing AI’s Benefits and Risks in Litigation and eDiscovery

Source: ComplexDiscovery