The proliferation of Internet technology has created numerous new markets as social coordination mechanisms, including those where human decision makers and computer algorithms interact. Because humans and computers differ in their capabilities to emit and process complex market signals, there is a need to understand the determinants of the provision of market information. We tackle the general research question from the perspective of new electronic credit markets. On online social lending platforms, loan applications typically contain detailed personal information of prospective borrowers next to hard facts, such as credit scores. We investigate whether a change of the market mechanism in the form of the introduction of an automated trading agent shifts the dynamics of information revelation from a high-effort norm to a low-effort information equilibrium. We test our hypothesis with a natural experiment on Smava.de and find strong support for our proposition.