Berkeley AI Policy Hackathon
This event has ended.
Entries
- View project: Recursive Fitness Alignment Protocol (RFAP)
Recursive Fitness Alignment Protocol (RFAP)
Team CC4CI — Caribbean Center for Collective Intelligence · St. John’s, Antigua
The Recursive Fitness Alignment Protocol (RFAP) is a minimal testbed for aligning reasoning systems—AI, human, or institutional—through recursive self-correction. Rather than optimizing for fixed goals or architectures, RFAP begins with the simplest fitness function (a random value between 0 and 1) and introduces a …
- View project: Round 1 Submission
Round 1 Submission
Team Rand#1 · San Francisco
Our round one submission for the UC Berkeley AI Policy Hackathon
- View project: Recommendation to Establish the California AI Accountability and Redress Act
Recommendation to Establish the California AI Accountability and Redress Act
Team Taylor Avenue · Cincinnati, OH
We recommend that California enact the California AI Accountability and Redress Act (CAARA) to address emerging harms from the widespread deployment of generative AI, particularly large language models (LLMs), in public and commercial systems. This policy would create the California AI Incident and Risk Registry …
- View project: CalSandbox and the CLEAR AI Act: A State-Led Vision for Responsible AI Governance
CalSandbox and the CLEAR AI Act: A State-Led Vision for Responsible AI Governance
Team Kalani & Gomes · Boston
The CLEAR AI Act proposes a balanced, forward-looking framework for AI governance in California that protects public safety without stifling innovation. Centered around CalSandbox, a supervised testbed for high-risk AI experimentation, and a Responsible AI Incentive Program, the Act empowers developers to build …
- View project: California Law for Ethical and Accountable Regulation of Artificial Intelligence (CLEAR AI) Act
California Law for Ethical and Accountable Regulation of Artificial Intelligence (CLEAR AI) Act
Team Kalani & Gomes · Boston
The CLEAR AI Act proposes a balanced, forward-looking framework for AI governance in California that protects public safety without stifling innovation. Centered around CalSandbox, a supervised testbed for high-risk AI experimentation, and a Responsible AI Incentive Program, the Act empowers developers to build …
- View project: Smart Governance, Safer Innovation: A California AI Sandbox with Guardrails
Smart Governance, Safer Innovation: A California AI Sandbox with Guardrails
Malaysia
California is home to global AI innovation, yet it also leads in regulatory experimentation with landmark bills like SB-1047 and SB-53. These laws establish a rigorous compliance regime for developers of powerful frontier models, including mandates for shutdown mechanisms, third-party audits, and whistle-blower …
- View project: FlexHEG Devices to Enable Implementation of AI IntelSat
FlexHEG Devices to Enable Implementation of AI IntelSat
Team Meerkats · Salt Lake City
The FlexHEG policy proposal combines Treasury Department tax credits and NSF technical assistance grants to create the hardware-enabled verification infrastructure necessary for implementing an IntelSat-like governance model for AI, ensuring both large corporations and smaller organizations can participate in a …
- View project: A Call for ‘Green AI’ in California: Evaluation of Existing and Alternative Regulations
A Call for ‘Green AI’ in California: Evaluation of Existing and Alternative Regulations
Team Green AI · Berkeley
The rapid growth of artificial intelligence, especially large-scale foundation models, poses critical environmental challenges. Training compute-heavy AI models consumes massive energy and water, leading to significant carbon emissions. California, a leader in climate and AI policy, doesn't have any legislation …
- View project: Data Trusts in AI Governance
Data Trusts in AI Governance
Team Athan & Tariq · Berkeley, CA
Data trusts are an alternative model for data governance that prioritise data subjects and creators in AI development.
- View project: From Sandbox to Standards: A Risk-Based Strategy for Responsible AI Innovation in California
From Sandbox to Standards: A Risk-Based Strategy for Responsible AI Innovation in California
Team Controlled Chaos Collective · Berkeley, California
To safeguard the interests of California and its population, it is time to create a comprehensive AI regulation system based on inclusion and experimentation to support the unfettered development of AI technology in our state. California has the unique opportunity to cultivate its own version of a prosperous “Good AI …
- View project: FlexHEG Devices to Enable Implementation of AI IntelSat
FlexHEG Devices to Enable Implementation of AI IntelSat
Team The Meerkats · Salt Lake City
The project proposes using Flexible Hardware Enabled Governance (FlexHEG) devices to support the IntelSat model for international AI governance. This framework aims to balance technological progress with responsible development by implementing tamper-evident hardware systems that enable reliable monitoring and …
Overview
As the home of Silicon Valley and a leader in AI governance discussions, California is at the forefront of regulation debates—especially with recent legislation like SB 53 and SB 1047. Supporters argue that AI regulation is essential for ethical use and to prevent bias and discrimination. Others worry that excessive oversight could stifle innovation and global competitiveness.Your challenge: Develop a bipartisan AI policy solution for California that fosters innovation while mitigating its most harmful risks.The purpose of this hackathon is to provide a space for students to advocate for solutions to AI policy issues that many lawmakers and technologists are thinking about, both in California and beyond. Over the past few years, the advancement of AI has presented unprecedented opportunities for innovation, as well as challenges for regulation. Through this event, participants can lend their perspective on what it takes for policymakers to weigh technological oversight and innovation and determine how they can coexist. Top teams will present their solutions to a curated panel of judges, with backgrounds ranging from academia, government, and industry.Everyone ages 18-25 who is interested in AI safety and policy is invited to participate. Our hackathon will be fully virtual, with co-working spaces for those based near Berkeley, CA. Due to the collaborative nature of policy making and ideation, all participants will enter in teams of 2. No prior experience in policy writing or AI safety is needed to participate and succeed.
Schedule
Our event will take place in two phases: -
Phase I:
Teams will work to submit a policy brief outlining their ideas and solution to our problem statement. This statement will be released on March 23, 2025 and policy brief submissions will be due on April 7, 2025. Finalists will be notified of their advancement by April 13, 2025.
Phase II:
Finalists will be tasked with developing their idea into a comprehensive policy memo that addresses the strengths, limitations, and stakeholders involved in their proposed solution. Examples of memos will be provided for those with no prior experience in policy writing. Submissions and presentations will occur during the weekend of April 26-27, 2025.
Upcoming Apart Sprints
All Sprints- Oct 23 - 25, 2026Online and in person
AI Collusion Research Sprint
A weekend research sprint on collusion between AI agents: when it emerges in markets and everyday workflows, how to detect and audit it, how it is carried, and what breaks it. Co-organized with Poseidon Research and AE Studio, online with in-person hubs at Collider in New York City and AI Safety Hong Kong. Top teams are invited to apply to the Apart Fellowship.
Read the brief: AI Collusion Research Sprint - Nov 13 - 15, 2026Online
AI x Epistemics Research Sprint
A weekend research sprint on AI for epistemics: evaluating whether models know how solid their claims are, building trust infrastructure that people and agents can consume, and shipping epistemic products that improve real decisions. Online, four tracks including an open track. Top teams are invited to apply to the Apart Fellowship.
Read the brief: AI x Epistemics Research Sprint