ÀÏÍõÂÛ̳

Join us on Monday, March 10, from 12:45–1:45 p.m. for a webinar featuring Jesmin Jahan Tithi as she introduces Z-Inspection, a methodology for accessing the trustworthiness of generative artificial intelligence systems.

Abstract

As generative AI (Gen-AI) continues to redefine the landscape of numerous industries, from drug discovery to creative arts, its integration into our socio-technical fabric necessitates a rigorous assessment of trustworthiness. Z-Inspection emerges as a pivotal methodology, offering a comprehensive framework to assess the trustworthiness of AI systems across their lifecycle. This process is grounded in the European Union’s High-Level Expert Group’s (EU HLEG) guidelines for trustworthy AI, yet exhibits the flexibility to adapt to diverse regulatory environments. This talk will delve into the details of Z-Inspection and its potential applications to Gen-AI. We will explore how this holistic approach facilitates identifying and examining ethical dilemmas, leveraging socio-technical scenarios crafted by interdisciplinary experts attuned to the specific use case.

The discussion will extend to the ethical quandaries inherent in Gen-AI, including bias amplification, the generation of deceptive content, privacy erosion, opacity, human skill diminution, employment disruption, and environmental impact. Our focus will be on preemptive analysis and the proactive cultivation of mitigation strategies through Z-Inspection, aiming to balance Gen-AI’s immense promise with the imperative of safeguarding ethical standards. Attendees will gain insights into operationalizing trustworthiness in AI development, ensuring that Gen-AI advancements are not only innovative but also aligned with societal values and ethical principles.