The paper addresses the ethical integration of artificial intelligence into philosophical counseling in Romania. We begin with the criteria used by Julia Clare and Richard Sivil to distinguish between practice and profession, and with the roadmap for recognizing the new profession of philosophical counseling in Romania, outlined in 2019 by Vasile Hațegan. Much of this process has already been accomplished; what remains is the cognitive dimension, which has yet to catch up with either online or algorithmically assisted practice. In the second part of the article, we draw upon Hațegan’s comparisons between philosophical counseling and psychotherapy, coaching, spiritual counseling, and mediation to identify five distinguishing criteria: counselor neutrality, the relational nature of the process, non-dependency, the non-therapeutic nature of the practice, and client autonomy. We treat these criteria as an ethical test for conversational AI agents. The APPAgpt platform proves particularly useful in this endeavor; through it, the American Philosophical Practitioners Association invites the public to converse with bots simulating dialogues with great thinkers, while simultaneously emphasizing the importance of the human counselor. The test yields nuanced results. When used by the counselor as a tool, artificial intelligence can operate within the boundaries of the practice; however, when placed in the counselor's role facing the client, it exceeds those limits. The article concludes with proposals for future occupational standards and the philosophical counselor's code of ethics: the inclusion of a competency unit dedicated to digitally mediated practice, as well as a decision-making procedure inspired by Henk van Luijk’s model.
