The AI startup Stability, known for developing the popular Stable Diffusion text-to-image model, is facing layoffs following the departure of its controversial CEO, Emad Mostaque.
An internal memo, which CNBC obtained, indicates that the company has made more than 20 employees redundant in an effort to “right-size” the business after experiencing a period of unsustainable growth.
In an email to staff on Wednesday night, Stability’s newly appointed co-CEOs, Shan Shan Wong and Christian Laforte, explained that the company must “restructure parts of the business, which will sadly mean saying goodbye to some colleagues. Those who are affected by this have been notified individually and we will be supporting them throughout this period.”
Wong and Laforte previously served as the chief operating officer and chief technology officer at Stability.
Stability AI has reduced its workforce by approximately 10%, impacting around 20 employees based on the company’s reported total headcount of 200, as per data available publicly.
According to a source who spoke to CNBC, the layoffs predominantly affect those in operational roles within the company. The employees have been informed about their redundancy, the source added, noting that they were not authorized to discuss the matter publicly.
Stable Audio 2.0
Stability’s latest release is called Stable Audio 2.0, which is an advancement to its text-to-audio AI model that can make music in less than a minute. The newer model is able to create music up to three minutes long with studio-grade audio quality at a resolution of 44.1 kHz in stereo.
Stable Audio 2.0 is also multimodal, meaning it can not only create audio tracks from simple text prompts but also from other audio files. Users can upload audio files and Stable Audio 2.0 generates fully produced samples using its audio-to-audio feature.
Stability AI has partnered with Audible Magic to integrate its Audio Content Recognition (ACR) technology, aiming to protect the intellectual property rights of music creators. This technology matches content generated in real time to help prevent copyright violations.
Additionally, the release of Stable Audio 2 has simplified the process of creating sound and audio effects from scratch. It now includes a broader range of options—from keyboard typing sounds to crowd noises, city ambiance, and more.
