Midjourney says it has banned employees from Stability AI, its rival company, accusing them of causing a disruption to the systems earlier this month during an attempt to extract data from its generative AI.
The AI company released an update on their Discord server on March 2nd, acknowledging that a prolonged server interruption was preventing generated images from appearing in user galleries.
In a summary of a meeting that took place on March 6th, Midjourney stated that behind this interruption was “botnet activity from paid accounts” (which the company specifically links to Stability AI employees).
Nick St. Pierre, a Twitter user who was able to hear the details of this meeting, said on the social network that Midjourney’s services were disrupted because “someone at Stability AI was trying to take over all the pairs of prompts and images in the middle of Saturday night.” St. Pierre also confirmed that Midjourney had linked several payment accounts to a person on the Stability AI data team.
Pierre directed his accusations at the CEO of Stability AI, Emad Mostaque, who responded on Twitter saying that he was investigating the situation and that Stability had not ordered such actions to be carried out.
“It is very confusing that two accounts have done this, and furthermore the team has not been scraping, as we have been using synthetic and other types of data, given that SD3 surpasses all other models,” said Mostaque, referring to the AI model Stable Diffusion 3, currently in preview phase. The executive stated that if the disruption was caused by a Stability employee, this would not have been intentional and “obviously it was not a DDoS attack”.
After this incident, Midjourney said on their Discord server that they have banned all Stability AI employees from using their service “indefinitely”. In addition, the company has implemented a new policy that will similarly prohibit access to employees of any company that engages in “aggressive automation” or causes service disruptions.
