Image generation AI Midjourney detects 'mass collection of prompts and images by BOT from Stable Diffusion developer' and permanently bans the account.

image-scraping



Midjourney , which develops and operates image generation AI, is reported to have indefinitely banned accounts owned by employees of Stability AI , which develops competing AI Stable Diffusion, from its services. Midjourney explains that Stability AI employees are suspected of using BOTs to perform data scraping that retrieves large numbers of prompt and image pairs.

Midjourney bans all Stability AI employees over alleged data scraping - The Verge
https://www.theverge.com/2024/3/11/24097495/midjourney-bans-stability-ai-employees-data-theft-outage



Image-scraping Midjourney bans rival AI firm for scraping images | Ars Technica
https://arstechnica.com/information-technology/2024/03/in-ironic-twist-midjourney-bans-rival-ai-firm-employees-for-scraping-its-image-data/



According to Midjourney user Nick St-Pierre, Midjourney detected BOT activity late at night on March 2, 2024.



Midjourney's meeting summary shared on Discord states, ``The service was suspended for 24 hours due to activity suspected to be BOT from a paid account,'' and ``The BOT in question was identified by a Stability AI employee using a combination of prompts and images. ``All Stability AI employees' accounts have been permanently banned from Midjourney services,'' and ``Actions that result in active automation or suspension of services are prohibited.'' The new policy 'will lead to the expulsion of all responsible employees of the company.'



According to St-Pierre, he used multiple paid accounts to scrape to obtain prompts and image URLs. The activity was so large and unusual that Midjourney was forced to temporarily suspend its service.

In response to St-Pierre's post, Stability AI CEO Emad Mostaq responded, ' What? ' and said, 'Stable Diffusion 3 outperforms all other models and uses synthetic and other data. We do not do any scraping because we do so.'



In addition, CEO Mostaak said, ``Although the behavior in question is not commendable, it is clearly not a DDoS attack and is not intentional,'' and said that he would conduct an internal investigation. In response, Midjourney founder David Holtz posted, ``We have sent information that will be helpful to our internal investigation.''



The Verge, an IT news site, requested comment from Midjourney and Stability AI, but as of the time of writing, neither company had responded.

in Software,   Web Service, Posted by log1i_yk