[Paper Review] ChatGPT: More than a Weapon of Mass Deception, Ethical challenges and responses from the Human-Centered Artificial Intelligence (HCAI) perspective

Alejo José G. Sisón, Marco Tulio Daza+2 more|arXiv (Cornell University)|Apr 6, 2023

Artificial Intelligence in Healthcare and Education11 citations

TL;DR

The paper analyzes ethical challenges of ChatGPT from a Human-Centered AI (HCAI) lens, reviews potential misuse as a weapon of mass deception, and proposes technical and non-technical mitigation strategies.

ABSTRACT

This article explores the ethical problems arising from the use of ChatGPT as a kind of generative AI and suggests responses based on the Human-Centered Artificial Intelligence (HCAI) framework. The HCAI framework is appropriate because it understands technology above all as a tool to empower, augment, and enhance human agency while referring to human wellbeing as a grand challenge, thus perfectly aligning itself with ethics, the science of human flourishing. Further, HCAI provides objectives, principles, procedures, and structures for reliable, safe, and trustworthy AI which we apply to our ChatGPT assessments. The main danger ChatGPT presents is the propensity to be used as a weapon of mass deception (WMD) and an enabler of criminal activities involving deceit. We review technical specifications to better comprehend its potentials and limitations. We then suggest both technical (watermarking, styleme, detectors, and fact-checkers) and non-technical measures (terms of use, transparency, educator considerations, HITL) to mitigate ChatGPT misuse or abuse and recommend best uses (creative writing, non-creative writing, teaching and learning). We conclude with considerations regarding the role of humans in ensuring the proper use of ChatGPT for individual and social wellbeing.

Motivation & Objective

• Frame ChatGPT within the Human-Centered AI (HCAI) framework to address ethics and wellbeing.
• Identify ethical risks of generative AI, especially deception and misuse.
• Recommend technical and non-technical measures to mitigate misuse and guide safe deployment.

Proposed method

• Provide an analytical description of generative AI concepts and core models (GANs, VAEs, BERT, GPT).
• Assess ChatGPT’s potentials and limitations to understand ethical implications.
• Review literature on deception, privacy, bias, and governance within the HCAI context.
• Propose a mix of technical safeguards (watermarking, detectors, fact-checkers) and non-technical measures (terms of use, transparency, HITL).
• Discuss best uses in education, writing, and learning to promote wellbeing.

Experimental results

Research questions

• RQ1 What ethical challenges does ChatGPT pose under the Human-Centered AI framework?
• RQ2 How do technical capabilities and limitations of ChatGPT influence its potential for misuse?
• RQ3 What mitigation strategies—technical and non-technical—can align ChatGPT with human wellbeing and trustworthy AI?
• RQ4 What governance and oversight mechanisms are warranted for reliable deployment of ChatGPT?

Key findings

• ChatGPT can function as a tool for deception due to its coherent, human-like outputs and low marginal cost of content creation.
• Misuse scenarios include academic deception, disinformation campaigns, and facilitation of criminal activities like malware generation.
• The model is not inherently safe or trustworthy; it hallucinates and lacks transparent, verifiable reasoning, challenging HCAI objectives.
• Authors propose technical measures (watermarking, detectors, fact-checkers) and non-technical measures (use terms, transparency, educator considerations, human-in-the-loop).
• HCAI emphasizes human agency, wellbeing, reliability, safety, and governance as guiding principles for responsible AI use.