Session
Holistic AI
Ella Shoup, Holistic AI, Private Sector, Western European and Others (WEOG)
Siddhant Chatterjee, Holistic AI, Private Sector, Asia-Pacific Group
Dr. Osman Gazi Gucluturk, Regulatory and Legal Lead at Holistic AI, Private Sector, Eastern Europe
Ella Shoup
Ella Shoup
12.6
16.10
16.3
16.b
17.14
17.17
Targets: 12: Ensure sustainable consumption and production patterns: The AI Act introduces stringent requirements for providers of high-risk AI systems and general-purpose AI models. Particularly regarding the GPAI models with systemic risk, the training and usage of which are energy intensive. Given that providers of these models, which are mainly large and translational companies, will be captured by the AI Act, understanding what the AI Act brings is certainly linked to this SDG target.
16: Promote peaceful and inclusive societies for sustainable development, provide access to justice for all and build effective, accountable and inclusive institutions at all levels: The AI Act introduces a set of comprehensive disclosure, reporting, and assessment requirements for operators engaging with high-risk AI systems or general-purpose AI models, the accountability aspect of which is important as these models could spread bias, deepen societal problems, and cause significant harm either as standalone systems or downstream applications developed based thereupon . More importantly, it contains substantial provisions against bias identification, assessment, and mitigation, which is aimed at reducing biased and discriminatory practices. Additionally, from an institutional perspective, it introduces a governance structure comprised of multi-stakeholder bodies, such as the advisory forum and scientific panel of experts.
Strengthen the means of implementation and revitalize the global partnership for sustainable development: AI Act is poised to be a global standard for AI regulation as the GDPR did in the field of data protection. It applies extraterritorially and references to international values.
The lightning talk will include 25 minutes of presentation by the speaker, which will include slides, followed by 5 minutes of Q&A from the audience.
Artificial Intelligence (AI) is in the process of reshaping the Internet landscape. The advent of Large Language Models (LLMs) and other advanced AI technologies has significantly altered our online interactions, content moderation practices, and approaches to digital safety and copyright concerns. As the integration of AI into our daily digital experiences deepens, both its benefits and its challenges have become more pronounced, leading to increased scrutiny from global regulators.
Among the most notable regulatory responses is the European Union (EU) 's forthcoming Artificial Intelligence Act (AI Act), which represents a significant step towards establishing comprehensive AI governance. Currently going through the final stages of the EU legislative procedure, this legislation aims to manage AI's integration across the EU by introducing obligations for high-risk AI (HRAI) systems and general-purpose AI (GPAI) models, setting standards that ensure technology's safe, secure, and rights-respecting deployment and comes with cross-sectoral as well as cross-border implications.
The EU AI Act is framed to be the AI governance equivalent of what the EU GDPR is to data protection, which is a global reference point and standard. Indeed, As the first comprehensive, binding AI governance framework in the world, it has significant potential to have a similar effect. However, its implementation demands extensive technical readiness, operational expenditure, and human resources, possibly causing a delay in its impact. Moreover, the high compliance costs could disproportionately benefit larger operators with ample financial resources, thereby creating a de-facto entry barrier in the AI ecosystem for smaller, less funded companies. Thus, while the AI Act holds the promise of setting a global standard for AI governance, its success and the equity of its impact will largely depend on addressing these critical challenges.
These uncertainties are poised to significantly impact the Internet, driven by the substantial overlap between major AI developers and leading online service providers and compounded by the increasing deployment of AI systems that transform how data is processed behind the scenes. Moreover, the AI Act introduces provisions that restrict, and in some cases prohibit, practices commonly seen on the internet, such as online advertising, which forms the financial backbone of many online services. This could lead to substantial shifts in how digital platforms operate and generate revenue.
In light of the above, this lightning talk provide a comprehensive analysis of the AI Act, paying a particular attention to its aspects affecting the Internet as we know. The talk will begin by examining the scope and objectives of the EU AI Act. Subsequently, we will assess its potential to reshape the Internet and address whether it can effectively balance innovation with the need for oversight. The discussion will also cover the AI Act’s broader implications, including its practical enforceability and its influence on global AI policy. Furthermore, we will critically analyse the challenges associated with the Act’s practical effect and the concerns that the AI Act might stifle AI innovation, considering both the risks and the necessary safeguards against them.
Prior to IGF, organizers will use the session’s page on the IGF website and social media channels to share informative and preparatory materials to help the audience better contextualize the topic.
Holistic AI will be taking questions during the Q&A from both the in-person and online audience. The online moderator will be active throughout the livestream to manage incoming questions and point the online participants to further resources.