Skip to main content

Discord is taking ChatGPT out of your browser and into your server

Discord is the latest brand to implement ChatGPT technology into its product as a push toward future AI functionality.

The server-based community plans to roll out a new feature to its Clyde chatbot next week, which will be available to a select number of servers with Discord alpha users. These testers will be able to use Clyde in an assistant fashion, having it start a thread for a group of your friends to hang out, recommend playlists, access GIFs, and more, according to a recent blog post.

Discord with the AI-powered Clyde chatbot.
Image used with permission by copyright holder

Discord’s announcement comes on the heels of SalesForce revealing how it will implement ChatGPT technology into its own Slack messaging platform. Snapchat also recently announced an AI chatbot called My AI for its app.

However, the Clyde chatbot inundated with AI has the potential to have the same issues and limitations as other products that have been using language models provided by ChatGPT’s parent company, OpenAI. The biggest example has been Microsoft’s Bing Chat, which released its AI chatbot in February. Users dealt with the depressed and unhinged responses generated from the chat in its early preview days, which led Microsoft to put restrictions on the turn limits in conversations. Those limits are now being lifted, apparently after some work on the brand’s part.

Discord is aware of the potential for its ChatGPT-based chatbot to go down a similar road and recommends that users not rely on Clyde for “advice, Discord support, or safety issues.” Users will also receive a warning when opting in to use the AI-updated Clyde for the first time.

Additionally, the company is aware that due to the knowledgeable demographic of its users, there is the potential for them to attempt jailbreaking the language models and force Clyde to say bad things or break terms of service. This is one of the main reasons the company has done a slow rollout starting with alpha users with whom it can ask for feedback, Anjney Midha, Discord’s platform ecosystem head, said in an interview with The Verge.

Clyde isn’t the only area where Discord is leveraging OpenAI’s technology. The company is also rolling out AutoMod AI, which leverages OpenAI technology to automatically moderate messages while understanding the context of the conversation. Discord is rolling this feature out to a “limited number of servers” now.

Discord is also adding AI-generated conversation summaries to catch you up on long threads, as well as Avatar Remix, which allows you to remix avatars for users in your server using generative AI.

Fionna Agomuoh
Fionna Agomuoh is a technology journalist with over a decade of experience writing about various consumer electronics topics…
A dangerous new jailbreak for AI chatbots was just discovered
the side of a Microsoft building

Microsoft has released more details about a troubling new generative AI jailbreak technique it has discovered, called "Skeleton Key." Using this prompt injection method, malicious users can effectively bypass a chatbot's safety guardrails, the security features that keeps ChatGPT from going full Taye.

Skeleton Key is an example of a prompt injection or prompt engineering attack. It's a multi-turn strategy designed to essentially convince an AI model to ignore its ingrained safety guardrails, "[causing] the system to violate its operators’ policies, make decisions unduly influenced by a user, or execute malicious instructions," Mark Russinovich, CTO of Microsoft Azure, wrote in the announcement.

Read more