You no longer need to sign up for an account before using the popular AI chatbot ChatGPT, following a change this week from OpenAI, the company behind the tool. The move results in access to a new, more limited version of ChatGPT and aims to broaden the tool's reach beyond its 100 million weekly users, as OpenAI battles rivals and stakes out territory on the AI ​​frontier. The change could lead to improvements to ChatGPT, but it also raises concerns.

The new version of ChatGPT is the same model that was previously available in the free tier – GPT-3.5 – but with some more restrictions, according to a company spokesperson. The change makes it easier to start conversations with the bot, the spokesperson said, adding that OpenAI's mission is to “make tools like ChatGPT widely available.”

That's a mission presumably shared by the companies behind rivals like Claude, Copilot, Firefly, and Gemini, who are also interested in convincing consumers to give their tools a try in the field without much, if any, oversight. As a result, this is a time when we are moving towards greater accessibility to generative AI — but also greater responsibility for users.

Andrew Frank, a renowned analyst at research firm Gartner, called this the “land grabbing phase of AI adoption.”

“At this stage, with any technology, the idea is to get as many users used to using your product as possible, and price is not an issue,” he said. “You really just want to establish the territory first and then figure out how to monetize it.”

Frank noted that OpenAI dropping its account requirements signals that “they perceive more of a competitive threat than they did a year ago when this was all so new.”

Starting April 1st, anyone will be able to visit chat.openai.com and start interacting with ChatGPT 3.5. OpenAI is rolling out access gradually, the company said in a blog post.

Logged in users will be able to do more with ChatGPT, like save and review chat history, share chats, and unlock features like voice conversations and personalized instructions. To set up an account, you must provide your email address, a password, your full name, and your date of birth — and agree to OpenAI's privacy policy.

ChatGPT also offers a subscription level. Through the $20 monthly ChatGPT Plus plan, consumers can access GPT-4, a more advanced model that can generate more text and respond to images, among other capabilities. (GPT-4 is also available for free through Microsoft Copilot.)

By eliminating the account requirement, OpenAI is throwing down a challenge to rivals — and positioning itself as a leader in adoption and potential market share, said Jason Alan Snyder, global chief technology officer at advertising agency Momentum Worldwide.

The advantages are numerous for both OpenAI and the public. To get started, you can immediately experience what ChatGPT has to offer. “This is great for the casually curious or anyone deciding if it's the right tool for them,” Snyder said.

It also invites what Snyder called “serendipitous inspiration” that can help you discover new uses for an AI chatbot. But it also benefits OpenAI by providing a larger, more diverse user base, which will help the company improve ChatGPT — and potentially lead to what Snyder called “more accurate and useful AI.”

By reducing friction for new users, OpenAI will gain an advantage in collecting data and feedback to be able to improve ChatGPT in the future. That's because OpenAI retains “certain data” of interactions, which it uses to improve its models (although users can opt out).

Regarding the account change, Rory Mir, associate director of community organizing at the Electronic Frontier Foundation, said any data minimization is a good thing, noting that privacy is a particular concern with AI chatbots. “We wouldn’t want Google to require accounts for searches, for example,” said Mir.

Meanwhile, as part of the goal of making accounts unnecessary, OpenAI is adding safeguards aimed at ensuring chatbot results are “appropriate for all age groups,” the spokesperson said. This includes blocking prompts and results across a wide range of categories.

Gartner's Frank said allowing someone to start using an AI chatbot is not without dangers. “We have certainly learned that anonymity on the internet has its risks,” he said. “Obviously, OpenAI recognizes the higher level of risk associated with allowing people to have anonymous access to a tool like this.”

But he questioned whether OpenAI is the right body to regulate itself.

As with “any other technical innovation that affects the entire public, there will need to be some type of neutral oversight that represents the public interest,” Frank said. “And I think maybe moves like opening OpenAI to everyone could accelerate the development of some kind of regulatory body.”

In addition to concerns about anonymous users creating mean-spirited content or scams, Snyder said the lack of accounts means ChatGPT can't remember user preferences, so it won't be able to provide a personalized experience for users who don't log in.

However, he is optimistic that removing barriers “unleashes a global, collaborative human-machine co-creation experience.”

Moving forward, said Snyder, we have to keep in mind that not all information generated by ChatGPT will be reliable or harmless, that even AI-assisted words and actions have consequences, and that how we use ChatGPT will affect how it evolves.

Editor's note: CNET used an AI engine to help create several dozen stories, which are labeled accordingly. For more, see our AI Policy.



Source