Global AI Safety Collaboration

Last month, I joined fellow Tech Global Institute members and advisors, Shahzeb Mahmood, Sheikh Waheed Baksh, Abdullah Safir, and Sabhanaz Rashid Diya in submitting our shared comments to White House Office of Science and Technology Policy following their request for input on the U.S. National AI Strategy. In this submission, we specifically draw on how AI governance can be advanced for low- and middle-income countries, who are disproportionately impacted, and the role of the U.S. in advancing international cooperation. Some key takeaways: 📌 AI ethics should be grounded in robust international human rights framework to close the global value alignment gap. Governance models building on ethics should be instituted through a more inclusive, representative and transparent process involving the Global Majority. 📌 Transparency is non-negotiable. However, it is critical that transparency efforts are substantiated by public consultations, reporting and independent audits, as well as updated on a regular frequency given the rapidly evolving nature of AI systems. 📌 The U.S. has a critical role in advancing labor protections for AI data workers in Global Majority regions through multilateral bodies, akin to similar protections offered to labor in manufacturing industry. 📌 AI models needs be culturally and socially situated including conducting impact and fairness assessment before global deployment. You can read the full submission here: https://lnkd.in/epHY7jXg. We welcome your feedback! #ArtificialIntelligence #GenerativeAI #GAI #GovTech

Big news from the UN this week- the Secretary-General António Guterres announced the creation of a new “High-Level Advisory Body on Artificial Intelligence” that’ll be dedicated to developing consensus around the risks and challenges posed by AI and how international cooperation can help make progress in meeting those challenges. Their mandate focuses on some key areas that are pretty critical to get right: - Global governance frameworks for AI use - Understanding risks like algorithmic bias and ensuring AI safety - Figuring out how to best leverage AI to accelerate the UN's Sustainable Development Goals - How to foster a globally inclusive approach while considering a multi-stakeholder approach The 39 member advisory body currently has a diverse representation from: - Government leaders, tech platforms, academia, and community - Voices from different countries like UAE, Kenya, Brazil, Pakistan and Singapore - Organizations like Mozilla, Google, Microsoft, Hugging Face, OpenAI Five of the members were recently named in TIME’s inaugural list of the 100 most influential people in AI: UAE minister of artificial intelligence Omar Al Olama, cognitive scientist Abeba Birhane, Google executive James Manyika, researcher and policy adviser Alondra Nelson, and computer science professor Yi Zeng. There are actually a few other international AI initiatives already in the works, including U.K. AI Safety Summit and G7 AI code of conduct. It’s interesting to see the UN take this proactive step to shape the responsible development of AI. I’m hoping this group creates space for nuance and various contexts in this global conversation on AI governance. I'm curious to see what guidelines, policies, outputs and conversations come out of their work.

🚨 Breaking News: Just Released! 🚨 The U.S. Artificial Intelligence Safety Institute (AISI) has unveiled its groundbreaking vision, mission, and strategic goals, released yesterday. This pivotal document sets the stage for the future of AI safety and innovation, presenting a comprehensive roadmap designed to ensure AI technologies benefit society while minimizing risks. Key Highlights: 🔹 Vision: AISI envisions a future where safe AI innovation enables a thriving world. The institute aims to harness AI's potential to accelerate scientific discovery, technological innovation, and economic growth, while addressing significant risks associated with powerful AI systems. 🔹 Mission: The mission is clear - beneficial AI depends on AI safety, and AI safety depends on science. AISI is dedicated to defining and advancing the science of AI safety, promoting trust and accelerating innovation through rigorous scientific research and standards. 🔹 Strategic Goals: 1. Advancing AI Safety Science: AISI will focus on empirical research, testing, and evaluation to develop practical safety solutions for advanced AI models, systems, and agents. 2. Developing and Disseminating AI Safety Practices: The institute plans to build and publish specific metrics, evaluation tools, and guidelines to assess and mitigate AI risks. 3. Supporting AI Safety Ecosystems: AISI aims to promote the adoption of safety guidelines and foster international cooperation to ensure global AI safety standards. 🔥 Hot Takes and Precedences: -"Safety breeds trust, and trust accelerates innovation." AISI's approach mirrors historical successes in other technologies, emphasizing safety as the cornerstone for unlocking AI's full potential. - Collaboration is Key: AISI will work with diverse stakeholders, including government agencies, international partners, and the private sector, to build a connected and resilient AI safety ecosystem. - Global Reach: By leading an inclusive, international network on AI safety, AISI underscores the necessity for globally adopted safety practices. This document is a must-read for anyone involved in the AI landscape. Stay informed and engaged as AISI leads the way towards a safer, more innovative future in AI. 🌍🔍 For more details, dive into the full document attached below. Follow WhitegloveAI for updates! #AISafety #Innovation #AIResearch #Technology #CLevelExecs #ArtificialIntelligence #AISI #BreakingNews #NIST Feel free to share your thoughts and join the conversation!

Last week, the UK and South Korean governments announced that 16 of the world’s leading #AI companies have agreed to a set of “Frontier AI Safety Commitments” during the “AI Seoul Summit.” The companies, including #Meta, #Google DeepMind, #OpenAI and China’s #Zhipu.ai, pledged to make sure advanced general-purpose AI systems would have adequate safety frameworks and guardrails to avoid misuse. In the worst cases, they would not deploy such a system if the risks cannot be kept below certain thresholds. It’s a great first step. Naturally, though, to avoid this being window-dressing and placating of regulators, defining the thresholds and enforcement mechanisms will be absolutely critical. And there are other questions and tensions to consider, as well: 1) Local vs. Global — Data, code, money and talent involved in AI flow and scale across borders.  Individual company and country pledges must transition to global integration and homogenization of safeguards, so the agreement mitigates any potential global loophole arbitrage. 2) Principles vs. Practice — The frameworks and thresholds need to be translated into practical, operationalized ethics and guidelines for doers creating these systems. 3) Proprietary vs. Open Source — The thresholds and enforcement mechanisms need to integrate the various issues posed by both closed and open systems, which require different safety and governance paradigms and approaches to the partner ecosystems of the large foundation model creators. (This is especially true given Meta’s LLAMA efforts and its press comments after the signing.) My colleagues and I at Cambrian Futures are encouraged by these commitments, and we hope the parties don’t skimp on the details and the definitions. Ultimately, a common set of strong guidelines can provide a springboard for greater innovation, because consumers and enterprise buyers see less risk and have greater trust to engage more fully with AI and share more data. Dan Zehr | Tobias Straube | Manu Kalia | Rehan Khan | Stephen Goodman | Michael Jeffrey | J. Nicholas Gross |  Julia Reinhardt |  Adam Tatarynowicz |  Burkhard Schrage |  Tom Sanderson |  Gladys Kuzmuk |  Kirthi Kumar |  Sana Pandey |  Lauren May Hildenbrand |  Ander Dobo | Ryan L. | David Babington