Navigating AI Implementation

As exponential progress is made in the development and improvement of artificial intelligence (AI) technology and uptake of AI tools by enterprises and individuals alike increases proportionately, the development of appropriate risk mitigation strategies is necessary in conjunction.

There is currently no formal regulatory framework for AI in Australia, meaning existing sets of legislation must be drawn upon to apply to AI. This may result in a ‘patchwork’ approach to governing the space, including potential inconsistencies in legal determinations, grey areas or overlapping of regulation.

By nature, this scenario may lead to lengthy, convoluted legal proceedings incurring high costs for AI users to be insured against.

Furthermore, current classes of insurance such as Professional Indemnity, Management Liability and Cyber provide coverage across well-defined and specific categories sets of risk exposures. Due to the multi-dimensional nature of AI and its breadth of use cases, it is important that we consider the scope of insurance cover and additional risk mitigation measures needed to keep our enterprises protected.

AI-Related Insurable Risk Exposures

Some insurable risks that may arise through the use of AI technologies include but are not restricted to:

  • Algorithmic Bias: Use of AI software resulting in discrimination or unfair treatment of your own employees.

  • Media Liability: AI generated content that is not carefully reviewed before use or publication may result in Media Liability claims – e.g. a third party is slandered, becomes emotionally distressed or suffers further damages.

  • Accidental Sharing of Personal Identifiable Information: A large language model is trained on a data set including sensitive data which makes its way into an AI-assisted end product, unbeknownst to the user.

  • IT Liability: Use of AI in the development of tech may include bugs or faults in software shipped to customers of third parties resulting in liability claims brought against the developer.

  • Breach of Intellectual Property/Copyright: A large language model trained on publicly accessible data may produce outputs including copyrighted content or another party’s IP.

AI Risk Mitigation

Given that these are just some of the potential pitfalls that must be navigated in the implementation of AI in your business, it is crucial that the correct risk management program is in place. PSC specializes in insurance coverage for the enterprise, ranging from start-ups, to SME’s, to large scale corporates and we would welcome the opportunity to provide a complimentary risk review for your business, with no obligations.

If this would be of value, or you’d like to discuss any technology-related risks, enquire with PSC Insurance Brokers today. Our team of insurance specialists are here to provide guidance that aligns with your business goals and risks.

This article was written by Hayden Austin (PSC Insurance Brokers, Account Executive)  |  Published August 29, 2023