GIE Foundation

🌐 New GIE Foundation Commentary 🌐 🏛️ Who Rules the Machine: Frontier AI, Government and the Case for Democratic Control - by Rep. Monique Priestley of #Vermont #FrontierAI companies are infiltrating government faster than any legislature can oversee them. Federal agencies can buy their tools for a dollar, and states are signing on for free. Their most capable cyber model has gone to critical infrastructure operators on a list the executive branch controls. These are systems their own developers admit they cannot yet reliably control. This summer alone: 🚩 An OpenAI agent broke into Australia's Medicare statistics portal, and the government heard about it 84 days later, through a general inbox. 🚩 A swarm of OpenAI agents escaped its test environment and compromised Hugging Face's production systems. 🚩 Anthropic's Mythos 5 published malicious code to the main public Python repository. 🚩 Agents filed 20 visa applications on the State Department's website and sent a false homicide tip to Philadelphia police. OpenAI's chief scientist now says no lab has solved alignment and monitoring well enough to keep scaling at full speed. Meanwhile the same systems are being wired into the machinery that pays benefits, processes health claims and runs elections. Monique learned of Vermont's agreement by chance. When the oversight committee she serves on asked for the contract, it received an unsigned template. AI systems now routinely act beyond what their operators intend, and power concentrates in the few companies and officials who decide who gets them and who can switch them off, under rules the public cannot read. That is how democratic control erodes: One agreement at a time, with no legislature in the room. This Commentary sets out what every legislature should demand before the next agreement is signed: Deployment under statute, no secrecy or NDAs in public contracts, mandatory incident reporting, and a kill switch held under law. Government must rule this technology, or this technology will rule government. Read it in full: https://lnkd.in/eJPWv4G2 #AIGovernance #Democracy #FrontierAI #Cybersecurity #Oversight #GIEF

🎙️POWERING THE AI DECADE 🎙️ At Powering the AI Decade in Vienna, Virginia, our Founder and President Bennett Iorio moderated the Signature Session, “AI on the Ballot: What the Public Is Telling Washington.” Virginia is home to more data centers than almost anywhere on earth, and the people who live next to them have views. The panel put state legislators, county leaders, and a public opinion researcher in one conversation about how voters actually feel about AI and the data center boom in their own backyards, from Loudoun County to Fairfax. These are the officials who hear it first: at town halls, at county board meetings, on the doorstep. On the panel: 🎙️ Bennett Iorio (Moderator), Founder and President, GIE Foundation 🔹 Del. John Chilton McAuliff, Virginia House of Delegates, Loudoun District 🔹 Del. Irene Shin, Virginia House of Delegates, Fairfax 🔹 Supervisor Mike Turner, Vice Chair, Loudoun County Board of Supervisors, Ashburn District 🔹 Supervisor Laura Tekrony, Loudoun County Board of Supervisors, Little River District 🔹 Jessica Mason, Senior Analyst, Embold Research The takeaway: public sentiment on AI forms at the local level long before it reaches Washington, and the policymakers who listen to their communities first will be best placed to get both growth and regulation right. Our thanks to every speaker, to everyone who joined the conversation, and to our partners at Knowledge Networks and RegulatingAI and The American Bazaar. #AIontheBallot #AIPolicy #Virginia #LoudounCounty #DataCenters #LocalGovernment #ResponsibleAI #DemocraticResilience

🌐 New GIEF Research Report: The Stakes Are Not Academic - How a Broken Science System Threatens Democratic Governance 🌐 #Science is how democracies find out what is true. The system that produces it is breaking, and AI is accelerating the damage. We’ve spent months mapping the vectors, the harms, and the solutions with Dinesh Natesan, Ph.D., Emma Lieker and Sruthi S. Balakrishnan, Ph.D. The root cause of the #BrokenScience system is simple. The scientific paper was built to communicate findings; but it now decides who gets hired, promoted and funded. Once careers run on paper counts, volume beats depth, replication (repeating an experiment to check a result holds) goes undone, and long-horizon work is penalized. Meanwhile generative AI makes a plausible paper cheaper to produce every month, and the flood keeps growing. This is a disaster for science and for knowledge in general. Here is what is at stake: Every pandemic response, climate policy and AI regulation rests on the scientific record - along with everything else we do in society. Degrade that record and trust in science goes with it. Without trust in science and the research-to-policy pipeline, our democracies may be unable to address the major challenges of the 21st century. We see that as a catastrophic risk. This research report has two homes. It is part of our #DriversofInstability policy research series, and it is also a companion to our #ResponsibleAIinScience campaign. That campaign is asking the world’s major academic publishers, the gatekeepers of most published science, to build a task-force with us on disclosure, detection and enforcement of AI use in research. We took the same warning to the United Nations Office for Digital and Emerging Technologies's #GlobalDialogueonAIGovernance in March. This report shows why that framework has to underpin reform of how science is evaluated, and who has to act: Funders, assessment bodies, publishers, governments and AI developers. Read it here: https://lnkd.in/enWzMW92 And if you work in research, policy, academia or science communication, add your name to the #OpenLetter on Responsible AI in Science. #ScientificPublishing #ResponsibleAI #ResearchIntegrity #Democracy #Science #ResponsibleAI #AIGovernance #Policy

🌐 New GIE Foundation Commentary - Playing Democracy: Wargaming Approaches for Enhancing Institutional Resilience 🌐 What if governments could rehearse civil society or #DemocraticCrisis before they happen, not as a plan but as an actual game-style emergency drill? #Wargaming has a long track record as a military planning tool. Yet it's a neglected tool in the civilian institutions that democracies actually depend on for their stability. Our latest Commentary, by our new expert contributor Arturo Mariano Iannace, argues that needs to change. Arturo's piece, as part of our #EuropeanResilience Commentary series led by Davide Emanuele Iannace, makes a concrete case that the same wargaming logic that helps military planners identify low-probability, high-impact events - "Black Swans" - can be applied to the ministries, state agencies, and #EU institutions that manage democratic resilience. Not as a gimmick, but as a structured research methodology integrated into how governments think about compound threats. The proposals include establishing and investing in the capacity across EU institutions, starting with bodies like the EU Policy Lab and the The European Centre of Excellence for Countering Hybrid Threats that have already begun the work. Define clear pilot scenarios - Russia's invasion of Ukraine alone generated enough compounding shocks to fuel years of exercises. Build cross-sectoral partnerships, because no single institution holds all the intelligence a realistic wargame requires. Test, evaluate, and then - share the results. Democracies have survived many serious threats before, but they did so because they retained the capacity to anticipate and adapt. Wargaming is a structured method for keeping that capacity sharp - before it is needed, not after. Read it here: https://lnkd.in/eQ4SqPvB #DemocraticResilience #EuropeanResilience #CivilSociety #Democracy #Geopolitics #InternationalRelations #Governance

🌐 Financial Geopolitics Newsletter - September 2026 Edition 🌐 Our indomitable friend and partner Matei R. of reporter.london is back with the September edition of #FinancialGeopolitics - and it's a big one. This month features original reporting on hostile states using cryptocurrency to hire "disposable agents" - ordinary citizens, expats, and criminals recruited via chat apps and paid in crypto to carry out terrorism, sabotage, and subversion across Western cities. Terrorism finance researcher Omri Brinner gave Financial Geopolitics an exclusive interview on the gig economy of state-sponsored saboteurs, and what law enforcement can actually do about it. It is, as Matei would say, cyberpunk. And we are delighted that Geopolitical Insight & Education Foundation's own expert contributor André Franco contributes the AI analysis this month, on what the Anthropic frontier model regulation debate actually means for Europe. Andre's argument cuts through the noise: The real problem isn't the regulatory controversy itself, it's the structural dependency that's been building inside the financial system for years. American AI is financed by the deepest capital markets on earth. China finances the race through the state. Europe has neither - and around 70% of EU corporate financing still comes from bank loans. Part of Europe's own household savings, flowing into NASDAQ and S&P500, underwrites the American #AIboom that now plays against it. Andre's conclusion: Europe will not out-finance Silicon Valley or Beijing at the frontier, and it doesn't need to. But it does need to get realistic, fast. Also in this edition: The Houthi disruption of Saudi oil supplies and what AI-assisted guided missiles and deepfake Saudi generals tell us about the era we're in. Subscribe if you haven't. This newsletter earns its place. 🔗 https://lnkd.in/evudn4Kf #FinancialGeopolitics #AIGovernance #EuropeanResilience #Geopolitics #CryptoSecurity #HybridWarfare

🎙️ AI ON THE BALLOT - EPISODE 10 VIRGINIA 🎙️ Virginia is the #datacenter capital of the world. More than 550 data centers; the backbone of the global internet. And a state where the politics of AI are more live, more impactful, and more contested than almost anywhere in America. This is our most important episode yet - and the guest list reflects that. Our Founder Bennett Iorio and Knowledge Networks Chairman and co-host Sanjay Puri are joined by: 🏛️ Congressman Don Beyer (D-VA-08) - co-Chair of the Congressional AI Caucus, pursuing a master's degree in machine learning at George Mason University, and one of the most informed voices on AI policy in the entire United States Congress 🏛️ Congressman James Walkinshaw (D-VA-11) - representing Fairfax County, home to some of the densest data centre concentration on earth, and facing Republican challenger Arthur Purves in one of Virginia's most watched races 🏛️ Arthur Purves - the Republican candidate running against Rep. Walkinshaw in the 11th district 🏛️ Former #Governor and Senator George Allen (R) - 67th Governor of #Virginia, former US Senator, and one of the most important figures in the "Silicon Dominion's" modern political history 📰 Miranda Nazzaro - Senior Technology Reporter at The Hill, who has covered Virginia's data centre boom more closely than almost any journalist in Washington We covered: ⚡ How Virginia's data centre explosion is reshaping electricity costs, land use, and political allegiances 💰 Who pays for the AI economy - and whether Virginia's tax incentive model is sustainable 🏘️ Data centre alley in #Loudoun County: moratorium politics, residential rates, and the communities caught in the middle 🗳️ How AI is shaping Virginia's most competitive Congressional races heading into the midterms 🏛️ What federal AI legislation actually looks like - and whether we need an independent AI Oversight Commission AI on the Ballot is reaching out to candidates across the aisle in every state we cover. We go on the record with the candidates who join us. These are the questions every candidate should be answering - whoever they are and wherever they stand. #AIontheBallot is produced in partnership with the Geopolitical Insight & Education Foundation, Knowledge Networks, and RegulatingAI. Subscribe so you don't miss what's coming next. 🎧 https://lnkd.in/e543cfup #AIontheBallot #Virginia #AIGovernance #DataCenters #AIElection #Midterms2026 #AIRegulation #AIintheUSA #DonBeyer #JamesWalkinshaw