Mitigating Values Debt in Generative AI: Responsible Engineering with Graph RAG
The rapid advancement of Generative AI technologies has unlocked unprecedented capabilities, transforming fields such as healthcare, education, and transportation. However, this progress often incurs a \textit{Values Debt}—ethical and operational deficits resulting from inadequate consideration of ethical principles during development. This paper explores the concept of Values Debt in Generative AI systems and introduces the Helpful, Honest, Harmless (HHH) framework as a practical solution to align AI systems with human values. By applying the HHH framework, we provide actionable recommendations for mitigating Values Debt and fostering responsible AI engineering. Through a case study of a chatbot developed using Graph RAG for aviation safety, we demonstrate how the HHH framework effectively addresses ethical challenges. These insights lay the groundwork for extending ethical AI practices across various domains.
Driven by a passion for responsible AI, I specialize in advancing methodologies and frameworks to operationalize ethical principles in AI systems. My research focuses on developing innovative approaches for evaluating and auditing AI technologies, ensuring core values such as fairness, trust, privacy, and security are deeply embedded in AI and machine learning systems.
With a commitment to rigorous ethical standards and value alignment, I aim to help organizations harness AI’s transformative potential while maintaining accountability and trustworthiness. I thrive on interdisciplinary collaboration, leveraging strong analytical and problem-solving skills to address complex challenges in AI and software engineering. My work is guided by the belief that responsible AI is not just a goal but an essential pathway to a more equitable and sustainable future.
Country: Australia Affiliation: CSIRO’s Data61 Research Interests: Responsible AI, AI Ethics, AI Auditing, Fairness in Machine Learning, Trustworthy AI, Privacy, and Security in AI Systems LinkedIn: https://www.linkedin.com/in/waqar-hussain/ X: https://x.com/Waqar_Husain_
Tue 29 AprDisplayed time zone: Eastern Time (US & Canada) change
09:00 - 10:30 | |||
09:00 10mDay opening | Opening Remarks RAIE Qinghua Lu Data61, CSIRO | ||
09:10 50mKeynote | Keynote 1 by Rick Kazman RAIE | ||
10:00 15mTalk | Insights and Current Gaps in Open-Source LLM Vulnerability Scanners: A Comparative Analysis RAIE Jonathan Brokman Fujitsu Research, Omer Hofman Fujitsu Research, Oren Rachmil Fujitsu Research, Inderjeet Singh Fujitsu Research, P: Vikas Pahuja Fujitsu Research, Aishvariya Priya Rathina Sabapathy Fujitsu Research, Amit Giloni Fujitsu Research, Roman Vainshtein Fujitsu Research, Hisashi Kojima Fujitsu Research Pre-print | ||
10:15 12mTalk | Mitigating Values Debt in Generative AI: Responsible Engineering with Graph RAG RAIE |