Microsoft-backed ChatGPT users can now surf the web, expanding the data the viral chatbot can access beyond its earlier September 2021 cutoff. The artificial intelligence startup said on Wednesday it had introduced an internet browser to the ChatGPT platform that could search the web for answers to queries and display them to users, bringing it closer to widely used AI voice assistants like Apple’s Siri. The company said the new feature is available for premium ChatGPT Plus and Enterprise users and will be expanded to free users soon.
The internet browser will use the Bing search engine to process users’ questions and deliver up-to-date information, OpenAI said in a post on its social media platform X (formerly Twitter). The chatbot is now “no longer limited by knowledge from before September 2021” and can answer complex queries such as who won the Oscars for best actor, soundtrack, or picture this year or what time the Formula 1 championship concluded last month, the company added. It also can provide current weather conditions, stock prices, and sports scores.
In the past, there have been privacy and ethical concerns about ChatGPT searching real-time information online because it could regurgitate harmful material, misinformation, or copyrighted content in response to queries. However, the company now has built-in safeguards informed by internal and external red teams to prevent such activities.
Having the ability to search the internet for answers could bring ChatGPT closer to rivals such as Google’s Bard and Amazon’s Alexa, both of which offer access to real-time information. It also makes the chatbot more useful in industries such as business, healthcare, and transportation, where the bot can find information about a particular topic or provide a map of a location, for example.
In addition to the browser, the ChatGPT update includes voice and image capabilities updates that will allow the bot to converse with users using their natural language or interact with images uploaded by them. It will also understand when users want to switch between written and spoken responses.
The new features are the latest in a series of enhancements to ChatGPT, which has drawn criticism from some people concerned it might be sifting through harmful material or misinformation to find an answer to their query. US regulators have also scrutinized the company for posing the risk of generating false information about consumers’ reputations. The chatbot has since vowed to improve its sourcing methods, and it is working with the Federal Trade Commission on how to address the issue. It has also said it will work to make the bot more trustworthy and transparent. The latest update to the software also enables it to take pictures of landmarks and provide details about them in response to specific questions. It will do this by automatically identifying and summarizing the most relevant online news articles.