The new features of Perplexity go beyond being just a chatbot.

Perplexity AI, known for its chatbot of the same name, has revealed that it will launch two new features offering greater flexibility in selecting sources of information: Internal Knowledge Search and Spaces. CEO Aravind Srinivas announced the availability of Internal Knowledge Search, a tool that allows users to search both the web and internal files of a team using multi-step reasoning and code execution. Previously, users could upload personal files for the AI to analyze, similar to how Gemini, ChatGPT, or Copilot works. With the new feature, Perplexity will now be able to explore personal documents and the web to formulate answers.

The company emphasizes that conducting searches that integrate internal and external data sources on a single knowledge platform will significantly improve businesses' productivity. Some initial users, such as financial services firms, have been using Internal Search to perform more thorough due diligence, accessing internal research and recent industry news.

On the other hand, Spaces are defined as an "AI-powered research and collaboration hub," allowing users to customize the Perplexity assistant to cater to specific data and needs of a project. Users will be able to select their preferred LLM model and determine how it should respond to their inputs.

The company assures that its Enterprise Pro clients will not have their private data used to train the AI, although standard Pro users must manually opt-in to this in the settings menu. The company highlights its commitment to data security and privacy and plans to soon integrate services like Crunchbase and FactSet for its Enterprise Pro clients.

These new features will initially be available to Perplexity Pro and Enterprise Pro subscribers, with a cost of $20 per month for the Pro plan and $40 per month per user for the Enterprise Pro plan. Despite its success with users, Perplexity has faced criticism from several websites whose news it uses as part of its training. In 2024, it was accused of “directly copying” articles from CNBC and Forbes, the latter of which has even threatened legal action. It has also received cease and desist orders from The New York Times and Conde Nast, and has been accused of plagiarism by Wired.