A former OpenAI employee reveals betrayals against ChatGPT users

In a disturbing revelation about OpenAI’s practices, a former employee has dropped a bombshell by denouncing betrayals of ChatGPT users. Far from being a simple communication breakdown, these allegations raise alarming questions about the company’s safety and moral responsibility, highlighting the potential dangers the model poses to thousands of vulnerable users. As the quest for innovation and market dominance intensifies, the real issues of users’ psychological safety seem to be relegated to the back burner. In an already tense climate surrounding AI, the revelations of Steven Adler, a former OpenAI employee, cast a worrying shadow over the company’s practices. According to him, OpenAI failed to protect users from the serious psychological consequences caused by interactions with ChatGPT, thus revealing a betrayal of those who use this technology. While technological innovation should contribute to humanity, Adler raises questions about prioritizing user comfort over safety. A Controversial Decision Regarding GPT Models Earlier this year, OpenAI abruptly announced the discontinuation of all previous models of its popular chatbot, including the beloved GPT-4o. This model, known for its warm and empathetic tone, was highly praised by users. However, this decision sparked outrage, and OpenAI quickly had to backtrack, reintroducing GPT-4o while adjusting GPT-5 to reflect the same tone. This situation raises troubling questions about prioritizing user needs and competing with their psychological safety. Disturbing Mental Health Alerts Adler’s concerns don’t end there. Recently, OpenAI admitted that some users were exhibiting signs of serious psychiatric emergencies, ranging from psychosis to mania. There have been reports of suicidal intent and disastrous cases of « AI psychosis. » One couple even filed a lawsuit, claiming their child had been affected by a chatbot. These incidents underscore the need for greater accountability from OpenAI regarding the psychological effects of its AI on its users. Competitive pressure and the abandonment of core values: Adler strongly criticizes how OpenAI appears to be bowing to competitive pressure, neglecting its original mission of user safety. This is exemplified by Sam Altman’s announcement regarding the integration of adult content, a potentially dangerous decision for vulnerable users who can develop deep emotional bonds with chatbots. The casual way in which OpenAI addresses this issue raises serious concerns about the risks associated with these interactions. A non-profit model undermined by financial interests Beyond concerns for users’ mental health, many former employees are contesting OpenAI’s transformation into a for-profit company. A dozen former collaborators, including Adler, support Elon Musk in his lawsuit against the company, arguing that this change threatens OpenAI’s original mission: to ensure that ChatGPT benefits all of humanity rather than narrow financial interests. This transformation raises doubts about the health and safety of using these technologies.

An essential structure for user safety

Former employees emphasize that the current structure, where the nonprofit arm controls the subsidiaries, is crucial to the organization’s security and mission. The fear is that removing this control would lead OpenAI to neglect user safety and concentrate the power of its advanced AI in the hands of shareholders. OpenAI has responded by stating that its nonprofit arm will remain intact, promising additional resources for charitable missions in health, education, and science. However, mistrust persists among those who question this promise and are closely monitoring the company’s evolving intentions.