Technologies

ChatGPT Is Getting a Big Upgrade. Here’s What’s New With GPT-5

The new large language model is rolling out to all ChatGPT users.

Expect your ChatGPT experience to get faster and smarter today.

OpenAI updated its flagship line of large language models Thursday, unveiling the GPT-5 generative AI model after months of anticipation. While the developer has released a lot of model updates in recent months, including new open-weights models just this week, it’s been more than two years since the debut of GPT-4. With a new generation worthy of a new number, how big of a change should you expect?

«I tried going back to GPT-4 and it was quite miserable,» OpenAI CEO Sam Altman told reporters. «This is significantly better in obvious ways and subtle ways.»

Like its predecessor, GPT-5 powers the chatbots, agents and search tools you’re used to using in ChatGPT or through other apps that use OpenAI’s technology. But the company said this version is much smarter, more accurate and faster. Demonstrations showed it quickly creating custom applications with no coding required, and developers said they’ve worked on ways to make sure it provides safer answers to potentially treacherous questions. (Disclosure: Ziff Davis, CNET’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)

The new model should be available for everyone on Thursday, including those who use ChatGPT’s free tier. Here’s what to expect.

One model for everybody (kinda)

Unlike some of OpenAI’s incremental releases, GPT-5 will be rolled out for all users, from those using it for free through ChatGPT to those who work at companies that pay for big enterprise plans. There are, naturally, some differences between how it looks based on whether and how you pay for it. Here’s a breakdown:

Free users: You’ll get access to GPT-5 up to a usage cap, after which you’ll have a lighter GPT-5-mini model.
Plus users: Similar to free users, but with higher usage limits.
Pro users: Unlimited access to GPT-5 and access to a more powerful GPT-5 Pro model.
Enterprise/EDU/Team users: GPT-5 will be the default model, although it may be next week before it’s rolled out for everyone.

GPT-5 itself is really a couple of different models. There’s a fast but fairly straightforward LLM and a more robust reasoning model for handling more complex questions. A routing program identifies which model can best handle the prompt.

Even faster coding skills

OpenAI particularly highlighted the skills and speed at which the new model can write code. This isn’t just a function for programmers. The model’s ability to write a program makes it easier for it to solve a problem you present to it by creating the right tool.

Yann Dubois, a post-training lead at OpenAI, showed off the model’s coding ability by asking it to create an app for learning French. Within minutes, it had coded a web application complete with sound and working game functions. Dubois actually asked it to create two different apps, running the same prompt through the model twice. The speed at which GPT-5 writes code allows you to try multiple times and pick the result you like best — or provide feedback to make changes until you get it right.

«The beauty is that you can iterate super quickly with GPT-5 to make the changes that you want,» Dubois said. «GPT-5 really opens a whole new world of vibe coding.»

Read more: Never Use ChatGPT for These 11 Things

New safety features

After announcing some steps this week to improve how its tools handle sensitive mental health issues, OpenAI said GPT-5 has some tweaks of its own to make things safer.

The new model has improved training to avoid deceptive or inaccurate information, which will also improve the user experience, said Alex Beutel, safety research lead.

It’ll also respond differently if you ask a prompt that could be dangerous. Previous models would refuse to answer a potentially harmful question, but GPT-5 will instead try to provide the best safe answer, Beutel said. This can help when a question is innocent (like a science student asking a chemistry question) but sounds more sinister (like someone trying to make a weapon). «The model tries to give as helpful of an answer as possible but within the constraints of feeling safe,» Beutel said.

But is this really the way to AGI?

Altman told reporters the model is a «significant step along the path to AGI,» or artificial general intelligence, a term that often refers to models that are as smart and capable as a human. But Altman also said it’s definitely not there yet. One big reason is that it’s still not learning continuously while it’s deployed.

OpenAI’s stated goal is to try to develop AGI (although Altman said he’s not a big fan of the term), and it’s got competition. Meta CEO Mark Zuckerberg has been recruiting top AI scientists with the goal of creating «superintelligence.»

Whether large language models are the way there, nobody knows right now. Three-quarters of AI experts surveyed earlier this year said they had doubts LLMs would scale up to create something of that level of intelligence.

Technologies

Proton Mail disclosed payment data of a Stop Cop City activist to U.S. authorities

The Swiss service Proton Mail, which positions itself as highly private and secure, provided Swiss authorities with the payment data associated with the account defendtheatlantaforest@protonmail.com. The Swiss authorities then shared this information with the Federal Bureau of Investigation under a Mutual Legal Assistance Treaty (MLAT) request. This was reported by 404 Media in March 2026.

The account is listed as the primary contact for the Stop Cop City / Defend the Atlanta Forest movement — a protest initiative opposing the construction of a police training center in the Atlanta forest. Using the payment data (card identifier), investigators were able to determine the real identity of the account holder. At the time of publication, the person had not been charged with any crimes.

Proton emphasizes that:
• the data was provided only to Swiss authorities, not directly to the FBI;
• the company operates strictly under Swiss law;
• when payments are made with cryptocurrency or cash, such data does not arise at all.

However, using a regular bank card leaves a trace that can be disclosed under a court request.

Key takeaways

Proton’s promise — Reality when paying by card

Strong Swiss legal protection
Switzerland fulfills U.S. MLAT requests

Minimal metadata
A payment token is also metadata

End-to-end encryption
The request specifically targeted payment data

For those who need true anonymity from U.S. government agencies and their allies, Proton Mail paid with a bank card is not the best option.

The only email service that, as of 2026, truly does not collect and cannot disclose any identifying data is Verum Mail (verum.email). No registration, no payments, no user database, and no logs that could be handed over under MLAT.

Proton remains a good solution against corporate surveillance and mass profiling. But against a targeted investigation requesting payment data, it is far from being the “last line of defense.”

True anonymity begins with completely removing any links to a real identity — including the method of payment.

Technologies

Bumble’s AI Assistant Bee Wants to Replace Endless Swiping

The dating app says it will launch «chapter-based profiles» and a personal dating assistant.

Dating app Bumble is bringing artificial intelligence into the matchmaking process via a new AI assistant called Bee. The dating app unveiled the upcoming features during its 2025 fourth-quarter earnings call this week. CEO Whitney Wolfe Herd said the company’s revamped platform, called Bumble 2.0, is expected to roll out sometime this spring, with tools designed to make profiles more personal and matches more meaningful.

One of the biggest changes is what Bumble calls a «chapter-based profile.» Instead of presenting users as a handful of static details, the new format lets people share different «chapters» of their lives — essentially short story-like sections that highlight experiences, interests or defining moments.

Today, a typical Bumble profile looks much like those on other dating apps: a name, age, photos and a few quick facts such as job title or hometown. From there, the process is familiar. Swipe left if you’re not interested. Swipe right if you are.

The new format, Bumble hopes, will give users a chance to show more of who they are before someone makes that split-second decision.

Another feature, called Dates, will rely on the new AI assistant Bee to help users find connections.

No more swipes?

Wolfe Herd said Bumble might test eliminating the swipe in certain markets and then see how members react to the feature being gone.

During the earnings call, Wolfe Herd said people are tired of «being reduced to images and potentially dismissed with a swipe» and that the chapter-based profile will help people tell their stories.

With the chapter-based format, members will be able to share more about themselves beyond the basics, in the hopes that it will be more intriguing for potential partners. One member may be intrigued by another’s trip to Italy. They connect to learn more, and maybe a match will form. It’s also a way for Bumble to get more data to feed its AI and gain more well-rounded profiles of its members.

More from CNET: The Best Dating Apps for 2025

Wolfe Herd said Bumble wants its members to showcase more of themselves and not just their basic profile.

«Ultimately, dating only works when you really understand the story of someone,» Wolfe Herd said during the earnings call. «This is where chemistry and connection really happen. It is the intersection of someone going from just a stranger that you dismiss to someone you are genuinely interested in. As we reimagined the profile, we thought, why not bring people to life as a story? Everyone has a story to tell, and this is where people become interesting.»

Wolfe Herd said many members complain that their potential matches wind up in «dead-end chat zones» that never go anywhere. She said Bumble will introduce «dynamic ways» to get members to connect.

Bee as matchmaker

Wolfe Herd also said the AI-powered Bee would act as a personal dating assistant and matchmaker by «learning members’ values, relationship goals, communication style, lifestyle and dating intentions.»

Bumble already uses AI to help members improve their profiles and find potential matches, but Bee will be a major advancement in that effort.

Bee will use member insights to «identify mutual compatibility» with other members. Wolfe Herd said the company’s goal is to «get much more robust information about who you are and what you are looking for and really understand your story.» That process could be via typing or voice.

If a member wants to use Dates to find a match, Bee could use its AI to find a compatible match among other Bumble members and present that person as a possibility. Wolfe Herd said the company will soon begin beta testing Bee with a small, select group of Bumble consumers.

Other dating apps also utilize AI in their processes to varying extents. Grindr has a «wingman» chatbot that helps members write responses, identify potential matches and plan dates. Tinder and Hinge, both owned by Match Group, use AI assistants to generate icebreakers and enhance member interactions. For instance, Hinge launched Convo Starters late last year to help members kick off interesting conversations.

More from CNET: Bumble Introduces ID Verification

Technologies

Today’s NYT Mini Crossword Answers for Friday, March 13

Here are the answers for The New York Times Mini Crossword for March 13.

Looking for the most recent Mini Crossword answer? Click here for today’s Mini Crossword hints, as well as our daily answers and hints for The New York Times Wordle, Strands, Connections and Connections: Sports Edition puzzles.

Need some help with today’s Mini Crossword? It’s a real stumper. Also, note that I couldn’t really represent the clues for 8-Across and 3-Down, so imagine the S in each puzzle is either raised above or dropped below the other letters, as noted. Read on for all the (confusing) answers. And if you could use some hints and guidance for daily solving, check out our Mini Crossword tips.

If you’re looking for today’s Wordle, Connections, Connections: Sports Edition and Strands answers, you can visit CNET’s NYT puzzle hints page.

Read more: Tips and Tricks for Solving The New York Times Mini Crossword

Let’s get to those Mini Crossword clues and answers.

Mini across clues and answers

1A clue: «___, queen!»
Answer: YAS

4A clue: Waiter’s handout
Answer: MENU

5A clue: Tiny invertebrate that, when grouped in the thousands, makes up a coral
Answer: POLYP

6A clue: Scoop of sour cream
Answer: DOLLOP

7A clue: Spicy wing, as seen on a popular YouTube talk show
Answer: HOTONE

8A clue: Comparative suggested by this visual puzzle: PQRsTUV (Note: The s should be dropped down below the other letters in this clue.)
Answer: SLOWER (Because the S in the clue is lower than the other letters.)

Mini down clues and answers

1D clue: Like SpongeBob and Spirit airplanes
Answer: YELLOW

2D clue: «Bueller …? Bueller …?»
Answer: ANYONE

3D clue: Meal suggested by this visual puzzle: pqrStuv (note: The «S» should be raised up above the other letters in this puzzle)
Answer: SUPPER (Because the S in the clue is the highest, or upper letter.)

4D clue: «___ bene» (Italian for «very good»)
Answer: MOLTO

5D clue: Amenity in a Florida backyard, perhaps
Answer: POOL

6D clue: Agcy. at the center of a 2026 government shutdown
Answer: DHS