Suno, currently embroiled in a lawsuit with the recording industry over using copyrighted songs to train its AI music model, has risen to become the fifth most-used generative AI service globally. Despite the legal challenges, the company is advancing its technology, launching its new V4 model for paid subscribers, which will soon be available to all users. This model offers more realistic music generation, featuring clearer production, improved vocals, and more surprising chord changes, making it desirable to listen to, according to co-founder Mikey Shulman. Suno operates from newly expanded offices near Harvard University and has grown its team from around 12 employees to over 50. Competing with AI giants like OpenAI, Suno distinguishes itself by focusing on aligning models with human tastes, utilizing user feedback to refine its technology. While AI-generated music has a distinct sound, V4 reduces this with more lifelike vocals and broader stereophonic effects.

Despite some industry resistance to AI-generated music, figures like Timbaland actively use Suno’s technology for creative projects. Suno prohibits the use of artists' names in its prompts to avoid potential voice reproduction issues. The platform has expanded features, allowing users to transform their own audio inputs into songs and inspiring a new generation of beatmakers. It's working towards a music-based user interface to engage both musicians and non-musicians. A new in-progress lyric-generating model is also available, providing more unique and human-like lyrics. Legal concerns about copyright infringement are recognized but not seen as insurmountable by Suno. Shulman envisions a collaborative future with labels and artists to build the next era of music.