Artificial intelligence is increasingly linked to productivity, growth, profitability, and competitiveness, with investors expecting firms to deploy AI solutions at scale. Firms therefore signal AI engagement through 10-K filings and the information contained in these filings can shape investors’ expectations about future value creation and attract investments. Yet AI information is difficult for investors to interpret because the underlying engagement is heterogenous and differs in technological maturity, uncertainty, and expected value. This dissertation examines how capital markets interpret different forms of AI engagement communicated in firms’ 10-K filings. The dissertation distinguishes AI engagement by stage, including realized adoption and intended planning, and by orientation, including explorative and exploitative engagement. Across two essays, it traces how AI information in 10-K filings is reflected in market prices and institutional portfolio decisions. The first essay examines two complementary pricing channels, event-time market reactions to newly disclosed AI information and the cross-sectional pricing of firm-level AI engagement. The second essay examines how transient and dedicated institutional investors adjust their ownership in response to information about different forms of AI engagement. Together, the essays contribution to our understanding of how investors infer the credibility, uncertainty, and value implications of different forms of AI engagement. Overall, this research has important implications for scholars examining the economic consequences of AI, firms communicating their AI initiatives, and investors evaluating AI-related disclosures.