Extended VAM for AI-Assisted Investment Among Indonesian Retail Investors

Many retail investors now use large language model (LLM) platforms like ChatGPT, Claude, and DeepSeek. These tools have changed how investment decisions are made. This study builds on the VAM model. This study tests how perceived benefits and sacrifices affect the perceived value of AI. We also explores how perceived value influences adoption intention. The model includes two benefit factors. These are perceived usefulness and perceived enjoyment. It also includes four sacrifice factors. These are technicality, perceived fee, security, and privacy. We use PLS-SEM to test seven hypotheses. The data come from a survey of 362 retail investors on the IDX. Two benefit factors were significant. Perceived usefulness had the strongest effect (β = 0.503, p < .001). Perceived enjoyment was also significant (β = 0.217, p = .011). Perceived value strongly influenced adoption intention (β = 0.939, p < .001). The four sacrifice factors were not significant. None passed the 0.05 threshold. Technicality even showed a positive effect. This was opposite to our expectation. The model explained 88.2% of the variance in perceived value. It also explained 93.2% of the variance in adoption intention. This study also extends VAM. Security and privacy are treated separately. Both are sacrifice factors. This difference is important in the AI investment context. The results of this study may be valuable for AI service providers, brokers, and the Indonesian regulators. It's a good way for them to boost their investment services with AI The aim is to establish more trust in AI and create a safer investment ecosystem. Keywords: VAM model, artificial intelligence, retail investors, Indonesia Stock Exchange, perceived value, security, privacy, large language model