Generative artificial intelligence (GenAI) tools, most visibly ChatGPT, have moved from a novelty to a routine input in how individual investors gather information, interpret news, and decide whether to buy or sell. This paper reviews recent empirical and experimental research on the role of GenAI in investor decision-making, drawing on studies published in accounting, finance, and behavioral economics outlets between 2015 and 2026. Three strands of evidence are examined. First, large language models show a measurable capacity to extract sentiment from financial text and to anticipate short-run stock price reactions, although this predictive edge appears to narrow as adoption spreads and markets adjust. Second, archival and survey data confirm that a substantial share of retail investors already rely on GenAI chatbots to screen and interpret company information, with usage shifting toward more sophisticated tasks as familiarity grows. Third, long-standing psychological barriers, particularly algorithm aversion and overconfidence, continue to limit how much weight investors place on machine-generated advice, even when that advice is demonstrably useful. The review also considers risks that are specific to generative systems, including fabricated outputs and an emerging “illusion of ability” that interactive tools can create in novice users. The paper closes with implications for financial educators, platform designers, and regulators, and it identifies gaps that future longitudinal and cross-market research should address.