Modern Trading in the Age of Artificial Intelligence: Examining Market Manipulation and the Case for Regulation

The rapid adoption of artificial intelligence (AI) powered trading systems has transformed modern stock exchange markets, including the New Zealand Exchange (NZX). AI-driven trading strategies offer significant benefits in liquidity, speed, and market efficiency. However, they also introduce new risks to market integrity, particularly AI-enabled market manipulation, emergent collusion, and distorted price discovery arising from system design, optimisation objectives, and opaque decision-making. Legal frameworks developed to address human-centred misconduct are increasingly strained by the scale, speed, and opacity of AI driven trading. This thesis examines whether New Zealand law can effectively regulate AI driven market manipulation on the NZX and, if not, how the legal framework must evolve. It analyses the adequacy of existing public law mechanisms, including the market manipulation provisions of the Financial Markets Conduct Act 2013, and assesses the extent to which concepts such as foreseeability, knowledge, and causation can be applied to AI-enabled trading systems.

The thesis also explores the role of private law remedies in promoting accountability and victim compensation. It assesses the applicability of negligence and statutory product liability under the Consumer Guarantees Act and Fair-Trading Act, identifying doctrinal constraints, particularly in cases involving pure economic loss, distributed agency, and the absence of a direct consumer relationship. In response, it proposes two targeted reforms: a statutory product liability regime under the FMC Act imposing strict liability for defective AI financial products, and a new tort of AI Governance Liability grounded in a statutory duty of diligent governance.

Ultimately, the thesis proposes an integrated legal and governance framework that combines preventive regulation with effective enforcement and compensation mechanisms. Key recommendations include pre-market approval for high-risk AI trading systems, enhanced transparency and explainability obligations, compliance by design requirements, continuous monitoring with tested kill switches, and clear board level accountability. It concludes that while New Zealand’s existing legal architecture provides a foundation for regulating AI-driven trading, targeted reforms are required to safeguard market integrity and maintain investor confidence in an increasingly automated financial ecosystem.