How Should AI Agents Avoid Losing User Trust When Providing Business Recommendations?

We have been delving deeply into the issues related to the field of artificial intelligence agents, and we hope to obtain practical feedback from those who are engaged in the development work in this area. As more and more users rely on agents to obtain purchase suggestions, tool recommendations, and service comparison information, agents are quietly becoming a new sales channel. However, in aspects such as the clear infrastructure and shared standards, this field still appears to be quite lacking in completeness: How should the transparency of information be maintained when agents recommend products or services? Should developers be able to obtain profits by providing truly useful recommendation services? Then, how should responsibility be attributed among recommendation, click, and conversion? And the most important question is - will any form of commercial operation automatically damage users' trust? We are currently conducting an investigation in this area, but the time is still relatively early. Therefore, we hope to obtain relevant information from developers and builders first. So, if you are developing artificial intelligence agents: Would you be willing to add commercial recommendation functions? What mechanism do you think is reasonable, transparent, and truly reliable? And what are your greatest concerns?