ElevenLabs Unveils Chat Mode to Expand Conversational AI Beyond Voice

ElevenLabs, the voice AI company known for its speech technology, has rolled out a new feature called Chat Mode, a text-only option for its conversational agents. Announced on August 19, the launch represents a significant shift for the company, which has until now primarily positioned itself as a leader in voice-first AI experiences.

Chat Mode allows developers and businesses to deploy conversational agents that interact entirely through text. The feature is aimed at situations where typing offers greater precision or comfort than speech, such as when users need to input sensitive information like email addresses or order numbers, or when speaking aloud is not practical. 

ElevenLabs said businesses can deploy a chat-based agent in just minutes through its SDKs, API, or even a single line of HTML, while existing voice agents can be switched to Chat Mode instantly. Earlier, the startup launched a new AI-based music generation service called Eleven Music. The feature launched into a legally fraught world of AI-generated music.

With this move, the three-year-old startup joins a growing list of AI companies venturing into music creation. However, the sector is under intense scrutiny, especially after major music labels recently sued startups like Suno and Udio for allegedly using copyrighted material to train their AI models.

To avoid such conflicts, ElevenLabs has reportedly partnered with Merlin Network, which represents independent music labels, and Kobalt Music Group, an independent music rights and publishing firm. These partnerships allow the company to train its model using licensed content and enable users to generate music that is cleared for commercial use.

Currently, around 20 customers, including creators and companies in TV, film, gaming, and wellness, have early access to the AI music tool. The company believes that AI-generated tracks could significantly cut costs for users in industries that rely on stock or production music.

Amidst this, the industry faces ongoing challenges. Organisations like the American Society of Composers, Authors and Publishers (ASCAP) have expressed concerns over the AI models training on existing music.