International AI Law Neumann 04232023.pptx

Current AI Landscape
Most Pressing Global AI Concerns
The Trouble with Global Collaboration
Regulation Examples to Consider
Biggest Challenges to International Law
The Independent Agency Problem
Possible Solutions
Conclusion
Research List

There are currently no global regulations or international laws being applied but the need is growing exponentially as AI technology explodes
The pace of technological changes and the unique challenges of things like independent agency creates new opportunities that are not addressed in existing international laws
AI that is developed in ways without oversight can negatively impact people through bias, exclusion, inaccurate identification, data misuse, and other negative risks
States and other organizations may be hesitant to overly regulate because of a fear of stifling innovation and economic growth
Now is the time to develop law, before there are larger problems, with economics and global stability in mind

1. Why Current International Regulations on Artificial Intelligence Are Insufficient Amy Neumann International Law 5550 April 23, 2023

2. Topics 2 Current AI Landscape Most Pressing Global AI Concerns The Trouble with Global Collaboration Regulation Examples to Consider Biggest Challenges to International Law The Independent Agency Problem Possible Solutions Conclusion R e search List

3. Current AI Landscape 3 AI is more and more available and accessible with increasing capabilities both good and bad Many organizations have expressed alarm (pause letters, ethics letters) Fears that ChatGPT , Bard (Google), CoPilot Microsoft), etc. are creating an “AI arms race” Concerns about risks or bias AI impacts many areas including employment, credit, weapons, privacy, policing, education, security, automation, loss of humanity, and more

4. Most P r essing Global AI Concerns 4 Existential risks (“genie out of the bottle”) Economic Inequity Ethical considerations Lack of transparency or “black box” technology Security and privacy "AI has the potential to facilitate both the protection and violation of human rights. A new legal framework is needed to ensure that AI is used to advance human rights, rather than undermine them." Jacob Turner, Robot Rules: Regulating Artificial Intelligence Automation leading to massive job losses Autonomous weapons Discrimination and bias Governance and regulation challenges Human rights

5. 5 5 AI Risk Levels: European Union Regulation Guidelines

6. R e gulation Examples to Consider 6 Organization Document Access European Union Artificial Intelligence Act https:// artificialintelligenceact.eu /the-act/ UN/UNESCO Ethics of Artificial Intelligence https:// www.unesco.org / en /artificial-intelligence/recommendation-ethics China (CAC) Numerous underway including by region, but China already uses social scoring banned by EU regulations https:// www.holisticai.com /blog/china-ai-regulation OECD Artificial Intelligence (AI) Principles https:// oecd.ai / en /ai-principles United States Pending, Accountability Act of 2022, AI BiIl of Rights, and state initiatives https:// www.whitehouse.gov /wp-content/uploads/2022/10/Blueprint-for-an-AI-Bill-of- Rights.pdf

7. 7 193 Countries Take First Global Steps 7 “ We see increased gender and ethnic bias, significant threats to privacy, dignity and agency, dangers of mass surveillance, and increased use of unreliable Artificial Intelligence technologies in law enforcement, to name a few. Until now, there were no universal standards to provide an answer to these issues .” UNESCO (United Nations, 2021)

8. The Trouble with Global Collaboration 8 Technology changes extremely fast Many existing human rights or other laws apply, for example the Universal Declaration of Human Rights, UN Guiding Principles on Business and Human Rights, and European Union's General Data Protection Regulation (GDPR) , but not everything has an example to follow AI crosses borders and can impact the entire world There are control, responsibility, accountability, and liability issues AI can become an independent entity or act in unknown or unforeseeable ways

9. Biggest Challenges to International Law 9 Different perspectives of human rights Fear of sharing proprietary information Biased data can create biased outcomes that are hard to see or predict but can impact billions Many players are participating including states, regions, international organizations, and non-governmental organizations (NGOs) No current overarching global standards or guidelines have been adopted by most There a perceived conflicts between oversight/regulation and economic/innovation freedom

10. The Independent Agency Problem 10 Much of existing law depends on “personhood” AI can suddenly perform in autonomous ways unknown to the designers AI can be created with or develop autonomy in decision-making, and can make wrong or biased decisions that can cause harm AI can become autonomous in making complex decision and how those decisions made might not be possible to know (credit scores, hiring, buying homes, criminal recidivism rates, etc.) Hard to determine ownership of ideas or content (intellectual property issues) Difficult to hold a non-human entity liable, especially with long or confusion chain of command or many players involved Weapons can make costly/lethal mistakes with or with a ”human in the loop” or cause international instability between countries

11. Possible Solutions 11 Develop additional groups to create guidelines with diverse types of people from many countries, backgrounds and industries to create frameworks applicable like UN treaties or similar Global algorithm auditing organizations can help control for misuse Existing international laws can be enforced where applicable New laws around independent agents (and their potential rights) could be created

12. Conclusion 12 There are currently no global regulations or international laws being applied but the need is growing exponentially as AI technology explodes The pace of technological changes and the unique challenges of things like independent agency creates new opportunities that are not addressed in existing international laws AI that is developed in ways without oversight can negatively impact people through bias, exclusion, inaccurate identification, data misuse, and other negative risks States and other organizations may be hesitant to overly regulate because of a fear of stifling innovation and economic growth Now is the time to develop law, before there are larger problems, with economics and global stability in mind

13. Thank You Amy Neumann 13

14. Research

15. Research