OpenAI has warned that users might form emotional attachments to its new ChatGPT voice interface, launched in late July. This concern is highlighted in a "system card" for GPT-4o, which details potential risks, safety testing methods, and strategies for mitigating emotional dependence.

OpenAI's GPT-4o, released in May, has been rated as posing a "medium risk" for influencing political opinions through text. The voice modality of GPT-4o, however, is considered to be low risk, according to a recent research paper.

"It's patently absurd that Apple isn't smart enough to make their own AI, yet is somehow capable of ensuring that OpenAI will protect your security & privacy!" Musk commented on X.

The announcement comes in the wake of increasing interest in voice-cloning technology, with several startups already offering similar solutions to the public. OpenAI's Voice Engine sets itself apart by claiming to replicate a person's voice with just 15 seconds of recorded speech. This breakthrough has sparked excitement among enthusiasts and industry professionals alike.

In response to Musk's legal action, OpenAI CEO Sam Altman resurfaced past interactions between himself and Musk, showcasing the latter's earlier support for the organization. However, tensions between Musk and OpenAI persist, with conflicting narratives emerging regarding their historical collaboration and disagreements over the company's strategic direction.

At the heart of the dispute is OpenAI's evolving relationship with Microsoft, which Musk says, that it diverges from the company's original mission of fostering public, open-source AI.