Whilst use of generative AI is growing rapidly, it does not mean that it is lawful or without risk. The consequences of implementing generative AI solutions into a business without addressing key risks could include:
- Sanctions/fines: The existing legislation (e.g. GDPR) and pending EU AI Act empowers regulators to raise considerable fines and sanctions for failure to assess and mitigate risk. The approach to sanction levels under the EU AI Act are proposed to be in excess of those under the GDPR.
- Litigation: Litigation in this area currently is focused on the training data aspect of generative AI solutions and misuse of publicly available data (e.g., copyrighted materials) to power the solutions. Monitoring in this area will increase and use of generative AI solutions powered by training datasets scraped from the internet will continue to carry risk of litigation.
- Wasted costs: Implementing a solution into the business which then needs to be unravelled could be costly (particularly where you have incorporated the solution into your own customer offerings). Beyond internal resource time this could result in unrecoverable development spend/wasted costs.
- Reputational damage: With the potential for IP infringement, confidentiality and data protection breaches coupled with the generation of unpredictable output, there is real risk of reputational damage if such solutions are implemented into the business without careful consideration.