Participants echoed the growing interest amongst business leaders to adopt GenAI tools and solutions to maximise efficiencies across enterprise functions. Participants also highlighted the socially beneficial potential of large language models in terms of enabling students, teachers and parents to access learning resources and farmers to understand complex export regulations in local languages.

Participants cautioned against the risks associated with the development and use of GenAI solutions, such as: proliferation of misinformation, violation of IP and privacy rights, identity frauds, and environmental damage. Participants also highlighted potential measures to circumvent these risks such as: conducting risk assessments, ensuring human oversight, implementing accountability protocols, ensuring transparency by informing users if output is AI-generated, and maintaining safety and security of AI systems.

Participants suggested a graded approach for integrating GenAI solutions in existing business workflows for companies with a low risk appetite.

Participants expressed concerns over the high costs of developing and deploying GenAI solutions. Building GenAI could be resource-intensive due to its dependence on cutting-edge infrastructure, high-quality datasets, extensive model training, and high carbon footprint. Participants suggested that enterprises could manage these costs by optimising the development of GenAI use cases and restricting access to them, planning staggered rollouts, reducing model parameters, and adjusting model accuracy according to the risk level of a given use case.

Participants stressed the need for upskilling and reskilling of the industrial workforce to make it GenAI-ready. The advent of GenAI is expected to disrupt the existing labour markets by automating certain jobs, while also altering the skills profiles for existing job opportunities to include complex problem-solving abilities, communication skills, and inter-disciplinary learning.