At ENSURE-6G’s 3rd event, “Advanced Wireless Network Security and Privacy,” Dr. Başak Ozan Özparlak delivered a crucial talk on “Regulatory frameworks for responsible AI in global 6G deployment.” Her presentation underscored the paramount importance of responsible AI in the forthcoming 6G networks, which are poised to be heavily AI-enabled [05:14]. Dr. Özparlak emphasized that “responsible AI” points to the human responsibility in governing these systems to ensure trustworthiness, rather than attributing liability to AI itself [04:52].
The EU’s Comprehensive Approach to Responsible AI
The EU has been at the forefront of establishing a robust regulatory landscape for AI:
- Ethical Foundations: Prior to legislation, the EU laid down ethical guidelines for trustworthy AI in 2019. These principles include human oversight, robustness (with a focus on cybersecurity), privacy, transparency, accountability, and fairness [07:32, 08:25].
- EU AI Act: This landmark act is the most comprehensive global legislation on AI, adopting a risk-based approach. Its broad scope, similar to GDPR, means it applies even to non-EU entities offering AI products or services within the EU market [13:56, 14:14]. It also complements GDPR regarding how AI processes data [16:36].
- Data-Related Regulations:
- GDPR: Remains foundational for data protection and privacy when AI processes data [16:12].
- Data Governance Act: Facilitates secure data sharing for AI services while upholding GDPR principles [09:24].
- Data Act: Particularly relevant for 6G, this act governs personal and non-personal data flow within the Internet of Things (IoT) ecosystem [09:50].
- Cybersecurity Regulations:
- NIS2 Directive: Crucial for 6G, this directive covers the telecommunications sector, mandating security by design, end-to-end encryption, and due diligence for supplier compliance [10:51, 17:19].
- Cyber Resilience Act: Promotes fundamental cybersecurity standards from the design phase for products with digital elements, including AI-embedded software, with full enforcement expected by 2027 [11:24, 11:46].
- Cybersecurity Act: Establishes a cybersecurity certification framework, essential for ensuring reliable and trustworthy products and services in 6G deployment [12:06].
- Other Key Directives: The Digital Markets Act and Digital Services Act aim to curb monopolistic structures in the digital market, including those involving AI [10:07]. The New Product Liability Directive now encompasses software and AI, introducing a presumption of defectiveness in technologically complex cases, which is highly significant for AI-driven 6G systems [19:16, 20:20]. The Radio Equipment Directive (RED) also ensures personal data and privacy protection for radio equipment, linking directly to 6G deployment [18:03].
Global Perspectives and Future Challenges
Dr. Özparlak also touched upon global efforts, citing the OECD’s updated ethical AI principles, which emphasize responsible stewardship, human rights, transparency, security, and accountability, while advocating for international cooperation and open-source AI development [22:17, 24:25].
Transparency remains a universally critical ethical consideration. While 100% transparency may be technically or legally infeasible, sufficient transparency is vital for users to understand when they are interacting with an AI system [25:30, 26:26].
Ongoing debates about regulating AI without stifling innovation, such as discussions in the US regarding state-level AI regulation, were also mentioned [29:06]. Ultimately, regulation is deemed necessary to ensure fair market conditions, security, and privacy, ensuring AI broadly serves society [31:17].
Looking ahead, Dr. Özparlak highlighted future regulatory issues for 6G:
- Spectrum Regulatory Innovation: Integrating non-terrestrial networks (NTNs) with terrestrial ones for 6G demands new approaches to spectrum licensing and assessing cybersecurity for space data processed by AI [32:13, 32:34].
- Digital Networks Act: This upcoming act, expected in Q4, will be crucial for 6G, potentially setting deadlines for copper networks and including major cybersecurity provisions for 5G/6G [33:59, 34:08].
In concluding, Dr. Özparlak underscored that given AI’s data-driven nature, regulatory frameworks prioritizing data privacy and security (such as GDPR and cybersecurity laws) are the most vital aspects to focus on for responsible AI in 6G [35:55, 36:04].
You can watch the full recorded talk here: Regulatory frameworks for responsible AI in global 6G deployment
