Leping

<I KAITSEVAGI Grant Agreement No. 0�4-1.'8/ J.,5/ l, Tartu 25.08.2025 The Board of Trustees of The Leland Stanford Junior University Frances C. Arrillaga Alumni Center 326 Galvez Street Stanford, California 94305-6105 Attention: Mr. Martin W. Shell Vice President, and Chief External Relations Officer Dear Members of the Board of Trustees: The Estonian Defence Forces ("EDF"), registry code 70008641, address Juhkentali 58, Tallinn 15007, represented by Brigadier General Viktor Kalnitski, Commandant of the Estonian Military Academy (EMA), pursuant to the statutes of the Estonian Defence Forces and a mandate issued pursuant to Directive No. 279 of the Commander of the Estonian Defence Forces of9 October 2014, is pleased to provide a contribution in the amount ofUS$200,000 to The Board of Trustees of the Leland Stanford Junior University ("Stanford"). This contribution shall be expendable and used to support The Future of Decision-making Project at the Center for International Security and Cooperation (CISAC) (or its successor) in the Freeman Spogli Institute for International Studies ("FSI") (the "Purpose"). :g 0 The acceptance of this contribution is based on these mutual understandings: 0 > � 1. Infrastructure Charge. The contribution will be subject to Stanford's infrastructure charge on gifts restricted to specific purposes at the time the gift funds are expended (as further described at https://adminguide.stanford.edu/chapter-3/subchapter-3/policy-3-3-1 or successor site). 2. Intellectual Property. All intellectual property or data resulting from the use ofthis contribution will be the property ofStanford. Stanford does not grant any rights in this contribution agreement to EDF with respect to any Stanford intellectual property or data. 3. Contingency. Ifthere comes a time when The Future ofDecision-making Project is no longer an active project at Stanford while any amount of the contribution remains unspent, then the remaining contribution funds shall be used under the direction ofthe director ofFSI for work and activities similar to or related to the Purpose. 4. Descriptive Proposal. The attached proposal ("Proposal") has been provided to EDF, and such Proposal for all components ofthis contribution is descriptive, rather than prescriptive, ofthe work that may be done under the Purpose and is not intended as a commitment to a specific line of inquiry, specific expenditure of time, specific expenditure of funds or specific research deliverable. Any narrative reports to EDF will reflect a general summary of the progress accomplished. Any financial reports will be summary reports offunds spent and need not include a line-item reconciliation with any proposed budget. 5. Staff and Budget. As the activity progresses, aspects of staffing and budget may change and the manner in which the activity is carried out may evolve while continuing to pursue the goals ofthe work. So long as the funds are being used for the general Purpose, EDF will allow considerable latitude in the specific expenditure offunds. 6. Academic Freedom. EDF recognizes Stanford's commitment to academic freedom and openness in research. With that in mind, EDF understands that this contribution does not provide EDF with real or perceived influence over the structure, process, content, or results of research, curriculum, programs, or other academic activities of Stanford. EDF and Stanford may collaborate regarding the use of the contribution, but EDF recognizes that Stanford will have the ultimate discretion in directing the use and administration ofthe contribution. 7. Name Use. Stanford and EDF may disclose factual information about the contribution in annual reports and to the extent required by law, regulation, rule or order, for tax purposes or as required to secure federal research funding. EDF may not issue a press release or other publicity statement in any medium regarding the contribution without the prior written consent ofStanford, which is provided on a limited basis. Any such approved statements must comply with Stanford's Name Use Guidelines (https://trademarks.stanford.edu/name-use-guidelines or successor site) and other Stanford policies. Except as otherwise provided in this paragraph, EDF will not, without Stanford's prior written consent, use (a) the name or quote of any Stanford faculty member employee or student in a Stanford capacity, (b) the name of any Stanford program or event, or (c) any Stanford trademark, service mark, trade name, symbol or other proprietary mark. 8. US Government Requirements. EDF understands that, if EDF is treated as a foreign source of funds under United States law, Stanford may need to disclose to the United States Government certain information about EDF, entities related to EDF or involved in making the contribution, and EDF's contribution to Stanford, including EDF's and related entities' names, addresses, and country of incorporation, as well as the date Stanford receives EDF's contribution, the amount, and other terms of the contribution, including possibly a copy of this agreement. 9. Miscellaneous. This agreement represents the entire agreement and understanding between Stanford and EDF with respect to the contribution described herein, and supersedes any prior and/or contemporaneous discussions, representations, or agreements, whether written or oral, of the parties regarding this subject matter. To the extent any provisions of this agreement conflict with any provision in the Proposal or any budget, if one was provided, the provisions of this agreement will control. This agreement may not be amended, altered or supplemented except by written agreement of the parties. This agreement may be executed in counterparts and all of these counterparts together shall be deemed to constitute one and the same agreement. This agreement may be executed by facsimile, .pdf or other electronic signature, and such facsimile, .pdf or other electronic signature shall constitute an original for all purposes. 2 Each party, by causing its duly authorized representative to sign below, acknowledges and agrees to the above terms. Electronically signed by: Martin W. Shell Reason: I have reviewed and am Martin W. Shell approving this document. Date: Aug 27, 2025 13:30:55 PDT Martin W. Shell ['1 Vice President, and Brigadier Chief External Relations Officer Command The Board of Trustees of the Leland Kaitsevae Akadeemia ulem Stanford Junior University 3  Attachment: Descriptive Proposal Proposal for The Future of Decision-making Project Center for International Security and Cooperation Stanford University Colin Kahl, PhD; Andrew Grotto, JD; Brad Boyd The Knowledge Gap-Artificial intelligence is changing how humans make decisions by automating the collection and processing of unprecedented amounts of data to provide predictions, insight, and recommendations. Our goal is to understand these potential impacts and to learn how governments can harness Al-enabled decision making in support of a peaceful, prosperous global order by examining decision dynamics in two domains: military decision-making and global public health decision-making. Al's impact on decision making will shape the global order. Unlike many previous revolutionary technologies such as electricity and computers, Al is not only improving human agency by accelerating human decision-making; it is also poised to supplant life or death decisions with machine intelligence. Al recommends targets for lethal force and turns intelligence data into operational recommendations in systems like the Israeli Defense Force's "Gospel" and could enable military systems to take battlefield actions without human oversight. Al is also being integrated into systems that help determine the effects of a pathogen, predict the pathogen's spread, and recommend responses. The rapid integration of Al into human decision-making promises many opportunities to optimize outcomes, but it also raises new risks. For example, integrating Al into decision-making may give nations advantages to deter aggression by predicting conflict and ensuring greater responsiveness, but it may also lead to runaway escalation as machines increase the speed of crises beyond human ability to control. In global health, Al may help contain pathogens before they become pandemics, but may also support development of novel bioengineered pathogens that do not exist in nature and are therefore more likely to generate surprise. There is so far little consensus on how to optimize hybrid human-machine decision-making structures because there is a significant knowledge gap on the potential effects that emerging Al-powered technologies will have on human organizations, processes, and even on individual psychology. Scholars have not yet illuminated the opportunities and risks that emerging technologies will provide for human decision-making. Nor have they provided an intellectual framework for managing change in society to ensure a human-centered approach. Novel Methodology-We propose an international, multi-disciplinary research project to understand how Al will affect human decision-making systems, identifying opportunities and risks of integrating emerging technology into human decision-making, and suggesting how human organizations and processes can evolve to achieve the best outcomes. We focus on military operational decision-making and global pathogen detection and response because both have the potential to avert or contribute to global human catastrophes in the future, and both have well-understood decision-making organizations employing deliberate processes that make detailed study and experimentation possible. Each also has a global community of expertise that will be harnessed for the broadest and most relevant perspectives, 4 illuminating different opportunities and risks while providing fertile ground for collaboration and insight among researchers. The project will have two phases over a minimum of two years. The first phase will be to understand decision-making systems in the two domains, document how Al technologies are currently being integrated, and characterize the resulting effects. We will focus on the implications of Al for a) the Napoleonic command structure commonly used in military decision-making and b) global public health surveillance. The first phase will culminate in a workshop showcasing the research from each case, and provide the start point for phase two. The second phase will build from the knowledge gained in phase one and suggest new decision-making models that could better optimize outcomes while mitigating risk. It will also suggest processes for managing the organizational, procedural, and psychological dynamics of new decision-making systems. This phase will focus on generating concepts and then testing those concepts against existing systems in virtual simulations conducted over a multi-day exercise. These simulations will compare proposed system performance against benchmark data assembled in phase one. Each phase will generate publication compilations and online resources for external researchers to review and contribute, including recommendations for policy-makers. Project Leadership-The project will be led by Dr. Colin Kahl, Andrew Grotto, and Brad Boyd. Dr. Kahl is the former Undersecretary of Defense for Policy in the U.S. Department of Defense and a Senior Fellow at the Center for International Security and Cooperation. Mr. Grotto is the Senior Director for the Program on Geopolitics, Technology, and Governance at Stanford University and the former Senior Director for Cybersecurity Policy for the Obama and Trump administrations. Mr. Boyd, currently a Visiting Fellow at the Hoover Institution, is the former Director of Al-enabled Warfighting System development at the U.S. Department of Defense, former National Security Advisor to U.S. Senator Angus King, and retired Colonel, U.S. Army. Collaboration-The future of decision-making project has already generated international interest from governments and universities in the United States, the United Kingdom, Australia, and Europe, all eager to collaborate with scholars at Stanford University because of its proximity to innovation and development in Silicon Valley, and its broad range of expertise in artificial intelligence and other technologies relevant to the research. We have already had informal exchanges and discussions with the British Ministry of Defence, the Danish Ministry of Defence, the Australian Department of Defence, the University of Exeter, the University of Adelaide, Cornell University, and the Massachusetts Institute of Technology. Other partners are being sought in East and South Asia. Policy Relevance-This research project will help maximize opportunities for human-centered outcomes while minimizing catastrophic risk by providing policymakers, government officials, and commercial developers with a theoretical framework and common frame of reference for improving decision outcomes. This research will also provide commercial developers with a body of research that will encourage the ethical development of new technologies and a more human focused approach for aligning powerful new technologies with the best interests of human society. A successful project will also provide a vehicle for continuing discussion about how to best support human-aligned decision-making over the next several decades. 5