Advances in Responsible AI Development and Governance

The field of artificial intelligence is moving towards a more responsible and transparent approach to development and deployment. Recent research has highlighted the importance of contextualizing AI evaluations, establishing standards and best practices for responsible AI, and enhancing trust through rigorous reporting and governance frameworks. There is a growing focus on the need for sociotechnical approaches to AI development, recognizing that technical and social factors are deeply intertwined. Bayesian statistics and reflexive prompt engineering are emerging as key tools for facilitating stakeholder participation and ensuring that AI systems are fair, transparent, and reliable. Furthermore, there is a increasing recognition of the need for holistic evaluation frameworks that integrate performance, fairness, and ethics, and for approaches to responsible governance that balance innovation and oversight. Notably, the papers 'Audit Cards: Contextualizing AI Evaluations' and 'Enhancing Trust Through Standards: A Comparative Risk-Impact Framework for Aligning ISO AI Standards with Global Ethical and Regulatory Contexts' propose innovative solutions for enhancing transparency and trust in AI reporting and governance. The paper 'Reflexive Prompt Engineering: A Framework for Responsible Prompt Engineering and Interaction Design' provides a comprehensive framework for responsible prompt engineering, highlighting the need for a delicate balance between technical precision and ethical consciousness.

Sources

Audit Cards: Contextualizing AI Evaluations

Framework, Standards, Applications and Best practices of Responsible AI : A Comprehensive Survey

Giving AI a voice: how does AI think it should be treated?

Existing Industry Practice for the EU AI Act's General-Purpose AI Code of Practice Safety and Security Measures

Position: Bayesian Statistics Facilitates Stakeholder Participation in Evaluation of Generative AI

Enhancing Trust Through Standards: A Comparative Risk-Impact Framework for Aligning ISO AI Standards with Global Ethical and Regulatory Contexts

Reflexive Prompt Engineering: A Framework for Responsible Prompt Engineering and Interaction Design

Evaluation Framework for AI Systems in "the Wild"

Approaches to Responsible Governance of GenAI in Organizations

The Malicious Technical Ecosystem: Exposing Limitations in Technical Governance of AI-Generated Non-Consensual Intimate Images of Adults

Built with on top of