California ACR 96 Asilomar AI Principles 2023

Summary

Expresses support by the California Legislature for the Asilomar AI Principles to guide the development of artificial intelligence and related public policy.

Supports creating AI with beneficial intelligence, ensuring robust systems that align with human values.

Encourages collaboration between AI researchers and policymakers and cooperation among AI developers to maintain safety standards.

Emphasizes AI safety, transparency in system failures and judicial use, and designer responsibility for AI implications.

Advocates for AI systems to respect human dignity, privacy, cultural diversity, and to empower people while sharing economic prosperity.

Stipulates human control over AI decision-making and discourages AI arms races.

Calls for cautious planning for advanced AI development, focusing on mitigating catastrophic risks.

Urges that superintelligence development serves ethical ideals and benefits all humanity.

Requests the Chief Clerk of the Assembly to distribute the resolution appropriately.

Summarizes the potential and importance of developing AI securely, reliably, and in alignment with human values.

WHEREAS, Over the last decade, artificial intelligence (AI) has demonstrated rapidly increasing competency across many fields, such as image recognition, speech recognition and translation, automated trading, autonomous vehicles, learning games from scratch, and the analysis of large datasets; and

WHEREAS, In the coming decades, AI is poised to disrupt many other domains previously serviced by human intelligence, including healthcare, law, finance, manufacturing, and education; and

WHEREAS, Further advancements in the application and performance of AI carry the potential to dramatically enhance individual and social well-being, so long as AI is developed in a manner that ensures security, reliability, and consonance with human values; and

Summarizes the Asilomar AI Principles discussed by experts for managing responsible AI development.

WHEREAS, In January 2017, AI researchers, economists, legal scholars, ethicists, and philosophers met at Asilomar, California, to discuss principles for managing the responsible development of AI; and

WHEREAS, The result of their collaboration was the 23 Asilomar AI Principles, which read as follows: Section I: Research Issues

Advocates for funding AI research on robustness, prosperity, legal alignment, and ethical status to ensure beneficial use.

(1) Research Goal: The goal of AI research should be to create not undirected intelligence, but beneficial intelligence.

(2) Research Funding: Investments in AI should be accompanied by funding for research on ensuring its beneficial use, including thorny questions in computer science, economics, law, ethics, and social studies, such as:

• How can we make future AI systems highly robust, so that they do what we want without malfunctioning or getting hacked?
• How can we grow our prosperity through automation while maintaining people’s resources and purpose?
• How can we update our legal systems to be more fair and efficient, to keep pace with AI, and to manage the risks associated with AI?
• What set of values should AI be aligned with, and what legal and ethical status should it have?

Encourages collaboration between AI researchers, policymakers, and developers to foster trust, transparency, and safety compliance.

(3) Science-Policy Link: There should be constructive and healthy exchange between AI researchers and policy-makers.

(4) Research Culture: A culture of cooperation, trust, and transparency should be fostered among researchers and developers of AI.

(5) Race Avoidance: Teams developing AI systems should actively cooperate to avoid corner-cutting on safety standards.

Requires AI systems to be safe, secure, and verifiably so where feasible.

Section II: Ethics and Values

(6) Safety: AI systems should be safe and secure throughout their operational lifetime, and verifiably so where applicable and feasible.

(7) Failure Transparency: If an AI system causes harm, it should be possible to ascertain why.

(8) Judicial Transparency: Any involvement by an autonomous system in judicial decision-making should provide a satisfactory explanation auditable by a competent human authority.

(9) Responsibility: Designers and builders of advanced AI systems are stakeholders in the moral implications of their use, misuse, and actions, with a responsibility and opportunity to shape those implications.

Specifies AI systems align with human values, and respect human dignity, rights, freedoms, and privacy.

(10) Value Alignment: Highly autonomous AI systems should be designed so that their goals and behaviors can be assured to align with human values throughout their operation.

(11) Human Values: AI systems should be designed and operated so as to be compatible with ideals of human dignity, rights, freedoms, and cultural diversity.

(12) Personal Privacy: People should have the right to access, manage, and control the data they generate, given AI systems’ power to analyze and utilize that data.

(13) Liberty and Privacy: The application of AI to personal data must not unreasonably curtail people’s real or perceived liberty.

Promotes AI benefits and prosperity for all, human decision control, societal respect, and avoiding AI arms race.

(14) Shared Benefit: AI technologies should benefit and empower as many people as possible.

(15) Shared Prosperity: The economic prosperity created by AI should be shared broadly, to benefit all of humanity.

(16) Human Control: Humans should choose how and whether to delegate decisions to AI systems, to accomplish human-chosen objectives.

(17) Non-subversion: The power conferred by control of highly advanced AI systems should respect and improve, rather than subvert, the social and civic processes on which the health of society depends.

(18) AI Arms Race: An arms race in lethal autonomous weapons should be avoided.

Advises against strong assumptions on AI capability limits; suggests careful planning for advanced AI potential.

Section III: Longer-Term Issues

(19) Capability Caution: There being no consensus, we should avoid strong assumptions regarding upper limits on future AI capabilities.

(20) Importance: Advanced AI could represent a profound change in the history of life on Earth, and should be planned for and managed with commensurate care and resources.

Requires risks from AI systems to undergo planning and mitigation efforts proportional to their impact.

(21) Risks: Risks posed by AI systems, especially catastrophic or existential risks, must be subject to planning and mitigation efforts commensurate with their expected impact.

(22) Recursive Self-Improvement: AI systems designed to recursively self-improve or self-replicate in a manner that could lead to rapidly increasing quality or quantity must be subject to strict safety and control measures.

(23) Common Good: Superintelligence should only be developed in the service of widely shared ethical ideals, and for the benefit of all humanity rather than one state or organization; and

Expresses support for the 23 Asilomar AI Principles as guiding values for AI development and policy.

WHEREAS, To date, the 23 Asilomar AI Principles have been signed by 1797 AI and robotics researchers and 3923 leaders in government, industry, and academia from across the globe; now, therefore, be it

Resolved by the Assembly of the State of California, the Senate thereof concurring, That the Legislature expresses its support for the 23 Asilomar AI Principles as guiding values for the development of artificial intelligence and of related public policy; and be it further

Resolved, That the Chief Clerk of the Assembly transmit copies of this resolution to the author for appropriate distribution.