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

iOS 26: AI Summaries Come Back to iPhone News Apps, but With a Warning

Apple initially disabled these summaries in January.

Apple released iOS 26 on Monday, a few months after the company announced it at the June Worldwide Developers Conference. The update brings a new Liquid Glass redesign, call screening and hidden features to your iPhone. The update also brings AI notification summaries for news and entertainment apps back to Apple Intelligence-enabled iPhone.

Apple disabled AI notification summaries for news and entertainment apps in January. That came a few weeks after the BBC pointed out in December that the feature twisted the media organization’s notifications and displayed inaccurate information.

Here’s what to know about those AI summaries and the new warning.

Don’t miss any of our unbiased tech content and lab-based reviews. Add CNET as a preferred Google source.

iOS 26 warns about summary inaccuracies

When I updated to iOS 26, I was greeted by some splash screens asking for various permissions. One splash screen was for the AI notification summaries. When you see this screen, you have two options: Choose Notifications to Summarize or Not Now. If you tap Not Now, the splash screen goes away.

If you tap Choose Notifications to Summarize, you’re taken to a new page where you’ll see three categories: News & Entertainment, Communication & Social and All Other Apps. Tapping one of these categories allows notification summaries for apps in that category. Beneath the News & Entertainment category, there’s a warning that gets outlined in red if you tap it.

«Summarization may change the meaning of the original headline,» the warning reads, adding, «Verify information.»

There’s also a warning across the bottom of the screen that reads, «This is a beta feature. Summaries may contain errors.»

After tapping the categories you want, tap Summarize Selected Notifications across the bottom of your screen. If you selected all the categories, this button will read Summarize All Notifications.

And if you don’t want these summaries, you can tap Do Not Summarize Notifications. If you allow these summaries and don’t like them, you can easily turn them off. Here’s how.

How to turn off AI notification summaries

1. Tap Settings.
2. Tap Notifications.
3. Tap Summarize Notifications.

4. Tap the Summarize Notifications toggle in the new menu.

You can also follow the above steps to turn AI notification summaries back on. You’ll have to select which categories you want these summaries for again, too.

For more on iOS 26, here’s my review of the OS, how to reduce the Liquid Glass effects in the update and how to enable call screening on your iPhone. You can also check out our iOS 26 cheat sheet.

Technologies

Amazon Prime Is Ending Shared Free Shipping. What to Know and When It Happens

How Prime Invitee program’s end could affect your free deliveries.

If you’ve been using someone else’s Amazon Prime membership for free shipping, but you don’t live in the same house, you may need to pay another subscription fee soon. According to Amazon’s updated customer service page, the online retail giant is ending its Prime Invitee benefit-sharing program Oct. 1.

Amazon’s Prime Invitee program is being replaced by Amazon Family, as reported earlier by The Verge. It includes many of the same benefits, but Amazon Family only works for up to two adults and four children living in the same «primary residential address» — a shared home.

You’ll still be able to use free shipping to send gifts elsewhere, but your Prime Invitees will no longer be able to use the perk.

Don’t miss any of our unbiased tech content and lab-based reviews. Add CNET as a preferred Google source.

Amazon isn’t the first company to prevent membership sharing between family and friends. The e-commerce giant is just the latest to follow Netflix’s account-sharing crackdown. While it’s unclear whether this change will work for Amazon, Netflix gained over 200,000 subscribers following its policy change. We also saw a similar account-sharing crackdown with Disney Plus and YouTube Premium.

Read more: More Than Just Free Shipping: Here Are 19 Underrated Amazon Prime Perks

What the Amazon Prime shipping crackdown means for you

If you’re the beneficiary of someone else’s Prime Invitee benefits, you have one more month to take advantage of the current program before the changes take effect.

Starting in October, you’ll have to get your own Amazon Prime subscription to benefit from the company’s free shipping program. First-time subscribers get a year of Prime membership for $15, but you’ll be stuck shelling out $15 a month to maintain your subscription thereafter.

Read more: Your Free Pass to Prime Day Deals (No Membership Required)

Why is Amazon ending the Prime Invitee program?

This move follows shortly after Reuters reported that Amazon’s Prime account signups slowed down recently despite an extended July Prime Day event. While the company reported blowout sales numbers, new Prime subscriptions didn’t meet internal expectations. In the US, they fell short of last year’s signup metrics.

According to Reuters, Amazon registered 5.4 million US signups over the 21-day run-up to the Prime Day event, around 116,000 fewer than during the same period in 2024, and 106,000 below the company’s own goal, a roughly 2% decline in both metrics.

By forcing separate households to have their own subscriptions, Amazon could be looking to attract more Prime accounts after previously failing to do so.

The new Amazon Family program (previously known as Amazon Household) offers Prime benefits to up to two adults and four children in a single home, including free shipping, Prime Video, Prime Reading and Amazon Music. The subscription also includes benefits for certain third-party companies, such as GrubHub.

Impulse Buys Under $25 on Amazon That Make Surprisingly Great Gifts

+17 More

See all photos

Technologies

Pokemon TCG Pocket’s Pack Points System Needs an Overhaul Yesterday

The pack-opening pity points system is pitiful. There’s a very easy way to improve it.

Pokemon TCG Pocket is more than a mobile game: It’s a money-making machine. The virtual trading card app raked in more than $900 million in its first six months, eclipsing even Pokemon Go’s revenue in the same post-release time span. As it turns out, fake Pokemon cards are just as much of a hot commodity as the real thing.

People love ripping open card packs, hunting down ones with their favorite illustrations of fan-favorite Pokemon. It feels great to beat the odds by pulling an elaborately-inked full art or a shiny secret rare. But it really starts to irk me when I’m missing only one or two cards from a set and I can’t get lucky enough to pull them out of a pack.

Pokemon TCG Pocket has a «pity points» system that’s supposed to make this feel less terrible: Every time you open a pack, you earn five pack points, which you can directly trade in for a card of your choosing.

You can trade in 35 points for a common card, but if you want to get the rarest cards from a set, they could eat up 500 points, 1,250 points or even a whopping 2,500 points each. That means you’d have to rip open 500 card packs in order to earn a single copy of one of Pokemon TCG Pocket’s rarest cards.

It sounds absurd (and it is), but that’s to be expected for a free-to-play game, especially one where the developer makes money by encouraging players to pay for extra card pulls. My real big issue with pack points is that they’re restricted to the expansion set you earned them in.

For example, I have 210 pack points for the latest card set, Secluded Springs, and I’ve been exclusively pulling those packs since it was released. I also have 700 pack points for the game’s first-ever expansion Genetic Apex — but those points are locked to Genetic Apex, and can’t be used for any other set. I’ve accrued hundreds of pack points, but they’re essentially useless to me because they won’t help me complete the sets I’m still missing cards in.

Pokemon TCG Pocket expansion sets are released on a monthly basis, which means no one really has time to earn enough pack points for a rare card before the next shiny slate of cards is dangled in front of your eyes. It propagates a desperate sense of FOMO that I’ve criticized in the past, but there’s a simple solution that would make the problem disappear overnight.

Instead of locking pack points to any one set, they should be an account-wide currency instead. Every time you earn pack points, they should be added to one large pool that you can use on any of the in-game card sets. That way, players wouldn’t have to feel a manufactured sense of guilt for ripping open packs from older sets.

While it’s customary for gacha games to have a pity system that guarantees a certain reward after a certain amount of pulls, it’s by no means a requirement for these games to have these systems. In a sense, I’m grateful that the pack points exist in Pokemon TCG Pocket in the first place.

I think we should always argue for a more consumer-friendly experience in modern gaming. Overhauling the pity system so that pack points can be used universally across all of the in-game card sets will make the game fairer and give more players a real chance to get the rarest cards.

It creates a greater sense of parity between free-to-play and paying players, and it might even cause some people to spend more money on pack openings to boot. Universal pack points are a win-win for players and DeNA alike.