Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
38 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
41 tokens/sec
o3 Pro
7 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Taxonomy to Regulation: A (Geo)Political Taxonomy for AI Risks and Regulatory Measures in the EU AI Act (2404.11476v1)

Published 17 Apr 2024 in cs.AI, cs.CY, and cs.LG

Abstract: Technological innovations have shown remarkable capabilities to benefit and harm society alike. AI constitutes a democratized sophisticated technology accessible to large parts of society, including malicious actors. This work proposes a taxonomy focusing on on (geo)political risks associated with AI. It identifies 12 risks in total divided into four categories: (1) Geopolitical Pressures, (2) Malicious Usage, (3) Environmental, Social, and Ethical Risks, and (4) Privacy and Trust Violations. Incorporating a regulatory side, this paper conducts a policy assessment of the EU AI Act. Adopted in March 2023, the landmark regulation has the potential to have a positive top-down impact concerning AI risk reduction but needs regulatory adjustments to mitigate risks more comprehensively. Regulatory exceptions for open-source models, excessively high parameters for the classification of GPAI models as a systemic risk, and the exclusion of systems designed exclusively for military purposes from the regulation's obligations leave room for future action.

PDF HTML Abstract
Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize File Document Download Save Streamline Icon: https://streamlinehq.com PDF File Document Download Save Streamline Icon: https://streamlinehq.com Markdown Bookmark Chat Bubble Oval Streamline Icon: https://streamlinehq.com Chat (Pro)
Definition Search Book Streamline Icon: https://streamlinehq.com
References (90)
  1. Frontier AI Regulation: Managing Emerging Risks to Public Safety. 2023. URL https://arxiv.org/abs/2307.03718.
  2. The Economist. AI Models Will Become Smaller and Faster. The Economist, 2023. ISSN 0013-0613. URL https://www.economist.com/the-world-ahead/2023/11/13/ai-models-will-become-smaller-and-faster.
  3. The White House. Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, 2023. URL https://www.whitehouse.gov/briefing-room/presidential-actions/2023/10/30/executive-order-on-the-safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence/.
  4. Dawn of the EU’s AI Act: Political agreement reached on world’s first comprehensive horizontal AI regulation | White & Case LLP, December 2023. URL https://www.whitecase.com/insight-alert/dawn-eus-ai-act-political-agreement-reached-worlds-first-comprehensive-horizontal-ai.
  5. Adrienne Héritier. The Accommodation of Diversity in European Policy-making and Its Outcomes: Regulatory Policy as a Patchwork. Journal of European Public Policy, 3(2):149–167, 1996. doi:10.1080/13501769608407026.
  6. Tanja A. Börzel. Pace Setting, Foot-Dragging and Fence-Sitting, Member States Responses to Europeanization. JCMS: Journal of Common Market Studies, 40(2):193–214, 2002. doi:https://doi.org/10.1111/1468-5965.00351.
  7. Clarifying the Costs for the EU’s AI Act, 2021. URL https://www.ceps.eu/clarifying-the-costs-for-the-eus-ai-act/.
  8. Jonathan Olsen. Public Policy and the Budget of the EU. In The European Union: Politics and Policies, pages 183–198. Routledge, New York, 7 edition, July 2020. ISBN 978-0-429-44072-4. doi:10.4324/9780429440724.
  9. The EU Policy Process. In Edoardo Ongaro and Sandra Van Thiel, editors, The Palgrave Handbook of Public Administration and Management in Europe, pages 805–821. Palgrave Macmillan UK, 2018. doi:10.1057/978-1-137-55269-3_42.
  10. Theories of the Policy Cycle. In Frank Fischer, Gerald Miller, and Mara Sidney, editors, Handbook of Public Policy Analysis, pages 69–88. Routledge, 2007. doi:10.4324/9781315093192.
  11. Council of the European Union. The Ordinary Legislative Procedure, 2024a. URL https://www.consilium.europa.eu/en/council-eu/decision-making/ordinary-legislative-procedure/.
  12. European Parliament. The Ordinary Legislative Procedure, 2024a. URL https://www.europarl.europa.eu/olp/en/ordinary-legislative-procedure/overview.
  13. European Parliament. Interinstitutional Negotiations, 2024b. URL https://www.europarl.europa.eu/olp/en/interinstitutional-negotiations.
  14. Anu Bradford. The Brussels Effect: How the European Union Rules the World. Oxford University Press, 2020. ISBN 9780190088583. doi:10.1093/oso/9780190088583.001.0001.
  15. The Brussels Effect and Artificial Intelligence: How EU regulation will impact the global AI market, 2022.
  16. Alex Engler. The EU AI Act Will Have Global Impact, but a Limited Brussels Effect, 2022. URL https://www.brookings.edu/articles/the-eu-ai-act-will-have-global-impact-but-a-limited-brussels-effect/.
  17. Risk Assessment at AGI Companies: A Review of Popular Risk Assessment Techniques from Other Safety-Critical Industries, 2023.
  18. Taxonomy of Risks posed by Language Models. In Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency, pages 214–229. Association for Computing Machinery, 2022. doi:10.1145/3531146.3533088.
  19. Sociotechnical Safety Evaluation of Generative AI Systems, 2023. URL https://arxiv.org/abs/2310.11986.
  20. Typology of Risks of Generative Text-to-Image Models. In Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society, pages 396–410. ACM, 2023. ISBN 9798400702310. doi:10.1145/3600211.3604722.
  21. Sam Clarke. Classifying sources of AI x-risk, 2022. URL https://forum.effectivealtruism.org/posts/e55QpEExmtkRjw9CD/classifying-sources-of-ai-x-risk.
  22. An Overview of Catastrophic AI Risks, 2023. URL https://arxiv.org/abs/2306.12001.
  23. TASRA: A Taxonomy and Analysis of Societal-Scale Risks from AI, June 2023. URL https://arxiv.org/abs/2306.06924.
  24. Stephen Clare. My model of how different AI risks fit together, 2024. URL https://forum.effectivealtruism.org/posts/Eg4PLFfZMcEck98vG/my-model-of-how-different-ai-risks-fit-together.
  25. Military Applications of Artificial Intelligence: Ethical Concerns in an Uncertain World. RAND Corporation, 2020. ISBN 978-1-977404-92-3. doi:10.7249/RR3139-1.
  26. Ioana Puscas. AI and International Security: Understanding the Risks and Paving the Path for Confidence-Building Measures, 2023. URL https://unidir.org/publication/ai-and-international-security-understanding-the-risks-and-paving-the-path-for-confidence-building-measures/.
  27. Rishi Iyengar. Who’s Winning the AI Race? It’s Not That Simple. Foreign Policy, 2023. URL https://foreignpolicy.com/2023/03/27/us-china-ai-competition-cooperation/.
  28. Will Knight. The Generative AI Boom Could Fuel a New International Arms Race. Wired, September 2023. ISSN 1059-1028. URL https://www.wired.com/story/fast-forward-generative-ai-could-fuel-a-new-international-arms-race/.
  29. A Comparative Analysis of the Definitions of Autonomous Weapons Systems. Science and Engineering Ethics, 28(5):37, 2022. ISSN 1471-5546. doi:10.1007/s11948-022-00392-3.
  30. Future of Life Institute. Autonomous Weapons Open Letter: AI & Robotics Researchers, 2016. URL https://futureoflife.org/open-letter/open-letter-autonomous-weapons-ai-robotics/.
  31. Michèle A. Flournoy. AI Is Already at War. Foreign Affairs, 102(6), 2023. ISSN 0015-7120. URL https://www.foreignaffairs.com/united-states/ai-already-war-flournoy.
  32. Terrorism and Autonomous Weapon Systems: Future Threat or Science Fiction?, 2023. URL https://cetas.turing.ac.uk/publications/terrorism-and-autonomous-weapon-systems-future-threat-or-science-fiction.
  33. The AI Power Paradox. Foreign Affairs, 102(5), 2023. ISSN 0015-7120. URL https://www.foreignaffairs.com/world/artificial-intelligence-power-paradox.
  34. The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation, 2018. URL https://arxiv.org/abs/1802.07228.
  35. Mustafa Suleyman. Containment for AI. Foreign Affairs, 2024. ISSN 0015-7120. URL https://www.foreignaffairs.com/world/containment-artificial-intelligence-mustafa-suleyman.
  36. House of Lords. Large language models and generative AI, 2024. URL https://publications.parliament.uk/pa/ld5804/ldselect/ldcomm/54/5402.htm.
  37. Ethical and social risks of harm from Language Models, 2021. URL https://arxiv.org/abs/2112.04359.
  38. Věra Jourová. Taylor Swift deepfakes nudge EU to get real about AI, 2024. URL https://www.politico.eu/article/europe-eye-fix-taylor-swift-nude-deepfake/.
  39. Brian Contreras. Tougher AI Policies Could Protect Taylor Swift–And Everyone Else–From Deepfakes. Scientific American, 2024. URL https://www.scientificamerican.com/article/tougher-ai-policies-could-protect-taylor-swift-and-everyone-else-from-deepfakes/.
  40. Artificial Intelligence’s Threat to Democracy. Foreign Affairs, 2024. ISSN 0015-7120. URL https://www.foreignaffairs.com/united-states/artificial-intelligences-threat-democracy.
  41. Risk in Focus: Generative A.I. and the 2024 Election Cycle, 2024. URL https://www.cisa.gov/resources-tools/resources/risk-focus-generative-ai-and-2024-election-cycle.
  42. Climate and conflict. Annual Review of Economics, 7(1):577–617, 2015. doi:10.1146/annurev-economics-080614-115430.
  43. Kristalina Georgieva. AI Will Transform the Global Economy. Let’s Make Sure It Benefits Humanity., 2024. URL https://www.imf.org/en/Blogs/Articles/2024/01/14/ai-will-transform-the-global-economy-lets-make-sure-it-benefits-humanity.
  44. Identifying and Mitigating the Security Risks of Generative AI. Foundations and Trends® in Privacy and Security, 6(1):1–52, 2024. ISSN 2474-1558, 2474-1566. doi:10.1561/3300000041.
  45. Geopolitical implications of AI and digital surveillance adoption, 2022. URL https://www.brookings.edu/articles/geopolitical-implications-of-ai-and-digital-surveillance-adoption/.
  46. The Repressive Power of Artificial Intelligence. Freedom on the Net 2023, Freedom House, 2023. URL https://freedomhouse.org/report/freedom-net/2023/repressive-power-artificial-intelligence.
  47. Google takes down Gemini AI image generator. Here’s what you need to know. Washington Post, 2024. ISSN 0190-8286. URL https://www.washingtonpost.com/technology/2024/02/22/google-gemini-ai-image-generation-pause/.
  48. Adrian Pecotic. Whoever Predicts the Future Will Win the AI Arms Race. Foreign Policy, 2019. URL https://foreignpolicy.com/2019/03/05/whoever-predicts-the-future-correctly-will-win-the-ai-arms-race-russia-china-united-states-artificial-intelligence-defense/.
  49. Rosanna Fanni. Why the EU must now tackle the risks posed by military AI, 2023. URL https://www.ceps.eu/why-the-eu-must-now-tackle-the-risks-posed-by-military-ai/.
  50. Lauren Kahn. Ground Rules for the Age of AI Warfare. Foreign Affairs, 2023. ISSN 0015-7120. URL https://www.foreignaffairs.com/world/ground-rules-age-ai-warfare.
  51. U.S. Department of Defense. Deputy Secretary of Defense Kathleen Hicks Announces Publication of Data, Analytics and AI, 2023. URL https://www.defense.gov/News/Releases/Release/Article/3577857/deputy-secretary-of-defense-kathleen-hicks-announces-publication-of-data-analyt/.
  52. U.S.-China Competition and Military AI, 2023. URL https://www.cnas.org/publications/reports/u-s-china-competition-and-military-ai.
  53. Sam Meacham. A Race to Extinction: How Great Power Competition Is Making Artificial Intelligence Existentially Dangerous, 2023. URL https://hir.harvard.edu/a-race-to-extinction-how-great-power-competition-is-making-artificial-intelligence-existentially-dangerous/.
  54. Safety not guaranteed: International strategic dynamics of risky technology races, 2022. URL https://www.governance.ai/research-paper/safety-not-guaranteed-international-strategic-dynamics-of-risky-technology-races. Working Paper. Centre for the Governance of AI.
  55. Heather M Roff. The frame problem: The AI “arms race” isn’t one. Bulletin of the Atomic Scientists, 75(3):95–98, 2019. doi:10.1080/00963402.2019.1604836.
  56. Artificial Intelligence Isn’t an Arms Race, December 2019. URL https://carnegieendowment.org/2019/12/11/artificial-intelligence-isn-t-arms-race-pub-80610.
  57. Paul Scharre. Debunking the AI Arms Race Theory. Texas National Security Review, 4(3):122–132, 2021. doi:10.26153/TSW/13985.
  58. Chris Meserole. Artificial intelligence and the security dilemma, 2018. URL https://www.brookings.edu/articles/artificial-intelligence-and-the-security-dilemma/.
  59. The Path to AI Arms Control. Foreign Affairs, 2023. ISSN 0015-7120. URL https://www.foreignaffairs.com/united-states/henry-kissinger-path-artificial-intelligence-arms-control.
  60. Kai-Fu Lee. The Third Revolution in Warfare. The Atlantic, September 2021. URL https://www.theatlantic.com/technology/archive/2021/09/i-weapons-are-third-revolution-warfare/620013/.
  61. MC Horowitz and P Scharre. Military uses of ai: A risk to international stability? AI and International Stability: Risks and Confidence-Building Measures, Center for a New American Security, pages 5–10, 2021. URL http://www.jstor.org/stable/resrep28649.5.
  62. Sam Biddle. OpenAI Quietly Deletes Ban on Using ChatGPT for “Military and Warfare”, 2024. URL https://theintercept.com/2024/01/12/open-ai-military-ban-chatgpt/.
  63. Jacob Stern. AI in Politics Is So Much Bigger Than Deepfakes. The Atlantic, 2024. URL https://www.theatlantic.com/technology/archive/2024/01/ai-elections-deepfakes-biden-robocall/677308/.
  64. World Economic Forum. Global Risk Report 2024, 2024. URL https://www.weforum.org/publications/global-risks-report-2024/.
  65. Tackling deepfakes in European policy. Publications Office of the European Union, 2021. doi:doi: 10.2861/325063.
  66. Majorities of Americans are concerned about the spread of AI deepfakes and propaganda | YouGov, 2023. URL https://today.yougov.com/technology/articles/46058-majorities-americans-are-concerned-about-spread-ai.
  67. Luminate. Bots versus ballots: Europeans fear AI threat to elections and lack of control over personal data, 2023. URL https://www.luminategroup.com/posts/news/bots-versus-ballots-europeans-fear-ai-threat-to-elections-and-lack-of-control-over-personal-data.
  68. New Hampshire investigating fake Biden robocall meant to discourage voters ahead of primary, 2024. URL https://apnews.com/article/new-hampshire-primary-biden-ai-deepfake-robocall-f3469ceb6dd613079092287994663db5.
  69. Is Argentina the First A.I. Election? The New York Times, November 2023. URL https://www.nytimes.com/2023/11/15/world/americas/argentina-election-ai-milei-massa.html.
  70. Valerie Wirtschafter. The impact of generative AI in a global election year, January 2024. URL https://www.brookings.edu/articles/the-impact-of-generative-ai-in-a-global-election-year/.
  71. Peter Conradi Bratislava. Was Slovakia election the first swung by deepfakes? The Times, 2024. ISSN 0140-0460. URL https://www.thetimes.co.uk/article/was-slovakia-election-the-first-swung-by-deepfakes-7t8dbfl9b.
  72. Todd C. Helmus. Artificial Intelligence, Deepfakes, and Disinformation: A Primer. RAND Corporation, July 2022. doi:10.7249/PEA1043-1.
  73. Europol. ChatGPT: The Impact of Large Language Models on Law Enforcement. Publications Office of the European Union, 2023. URL https://data.europa.eu/doi/10.2813/255453.
  74. How persuasive is AI-generated propaganda? PNAS Nexus, 3(2), 2024. doi:10.1093/pnasnexus/pgae034.
  75. Comparing the willingness to share for human-generated vs. ai-generated fake news, 2024. URL https://arxiv.org/abs/2402.07395.
  76. European Commission. Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL LAYING DOWN HARMONISED RULES ON ARTIFICIAL INTELLIGENCE (ARTIFICIAL INTELLIGENCE ACT) AND AMENDING CERTAIN UNION LEGISLATIVE ACTS, 2021. URL https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:52021PC0206.
  77. High-Level Expert Group on AI. Ethics guidelines for trustworthy AI | Shaping Europe’s digital future, 2019. URL https://ec.europa.eu/newsroom/dae/document.cfm?doc_id=60419.
  78. European Commission. White Paper On Artificial Intelligence - A European approach to excellence and trust, 2020. URL https://commission.europa.eu/document/d2ec4039-c5be-423a-81ef-b9e44e79825b_en.
  79. Kai Zenner. The AI Act: All publicly available documents in March 2024, 2024. URL https://www.kaizenner.eu/post/aiact-part3.
  80. Council of the European Union. Proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (Artificial Intelligence Act) and amending certain Union legislative acts - Analysis of the final compromise text with a view to agreement, 2024b. URL https://data.consilium.europa.eu/doc/document/ST-5662-2024-INIT/en/pdf.
  81. Gian Volpicelli. EU set to allow draconian use of facial recognition tech, say lawmakers, 2024. URL https://www.politico.eu/article/eu-ai-facial-recognition-tech-act-late-tweaks-attack-civil-rights-key-lawmaker-hahn-warns/.
  82. European Commission. European AI Office, 2024. URL https://digital-strategy.ec.europa.eu/en/policies/ai-office.
  83. General Purpose AI Systems in the AI Act: Trying to Fit a Square Peg Into a Round Hole. We Robot 2023, 2023. doi:10.2139/ssrn.4644701. URL https://www.ssrn.com/abstract=4644701.
  84. Elizabeth Seger. What Do We Mean When We Talk About "AI Democratisation"?, 2023. URL https://www.governance.ai/post/what-do-we-mean-when-we-talk-about-ai-democratisation.
  85. Alex Friedland. What Are Generative AI, Large Language Models, and Foundation Models?, 2023. URL https://cset.georgetown.edu/article/what-are-generative-ai-large-language-models-and-foundation-models/.
  86. Théophane Hartmann. Behind France’s stance against regulating powerful AI models, 2023. URL https://www.euractiv.com/section/artificial-intelligence/news/behind-frances-stance-against-regulating-powerful-ai-models/.
  87. European Commission. Artificial Intelligence - Questions and Answers, 2023. URL https://ec.europa.eu/commission/presscorner/detail/en/qanda_21_1683.
  88. David Evan Harris. Open-Source AI Is Uniquely Dangerous - IEEE Spectrum, 2024. URL https://spectrum.ieee.org/open-source-ai-2666932122.
  89. Carl Franzen. Mistral shocks AI community as latest open source model eclipses GPT-3.5 performance, 2023. URL https://venturebeat.com/ai/mistral-shocks-ai-community-as-latest-open-source-model-eclipses-gpt-3-5-performance/.
  90. OpenAI. ChatGPT, 2024. URL https://chat.openai.com.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (1)
  1. Sinan Arda (1 paper)