Feed aggregator

OpenAI's latest AI model GPT-5.2 is here. But how does it compare to its biggest competitor, Google's Gemini 3?

The ChatGPT creator launched the GPT-5.2 on Thursday, and it's currently rolling out to paid ChatGPT customers, with wider access to follow. It comes on the heels of a leaked memo, in which OpenAI CEO Sam Altman declared "code red" for the AI company. OpenAI is facing steeper competition from Google after the release of its latest AI model, Gemini 3. Google's "most intelligent" AI model yet was surging past its AI competitors in various benchmark tests and on the AI leaderboards at LMArena. 

In addition to that, a post from former Googler Deedy Das claimed that in the two weeks since Gemini 3 had launched, OpenAI lost nearly 6 percent of its visitors.

SEE ALSO: How to try GPT-5.2, the new ChatGPT model series from OpenAI

So, how does OpenAI's GPT-5.2 stack up against Google's Gemini 3? So far, the models seem evenly matched on benchmark performance, pricing, and other comparisons, with both models gaining an edge in specific areas.

First, it should be noted that GPT-5.2 just came out, and we'll know more in the days ahead. Unsurprisingly, OpenAI claims that GPT-5.2 excels over other AI models when it comes to "professional knowledge work." OpenAI describes this as GPT-5.2 being "better at creating spreadsheets, building presentations, writing code, perceiving images, understanding long contexts, using tools, and handling complex, multi-step projects."

GPT-5.2 vs. Gemini 3: LMArena leaderboards

Based on the more neutral LMArena leaderboards, GPT-5.2 appears to be impressing users in at least one key area: web development.

The leaderboard has the higher-end GPT-5.2-high in second place, behind Claude Opus 4.5. Gemini 3 Pro is in the fourth spot, and the base GPT-5.2 model is in sixth. So, for web development tasks, it appears GPT-5.2 excels above Gemini 3. Though, again, these are early days.

So far, GPT-5.2 is unranked on the rest of LMArena's leaderboards.

However, GPT-5.2 has a steep hill to climb. Various versions of Gemini 3 currently top the leaderboards for text, vision, text to image, image edit, and search, with the closely related Google Veo 3 models topping the remaining text to video and image to video categories. Gemini 3 also tops the overall leaderboard, where GPT-5.2 remains unranked.

GPT-5.1-high is currently ranked sixth in the overall rankings, which explains why Altman might be seeing red.

GPT-5.2 vs. Gemini 3: Benchmark tests

At this time, we can't independently verify the benchmark results OpenAI has reported for GPT-5.2, and third-party rankings from companies like ScaleAI so far don't include the new model. However, looking at the benchmark results reported by Google and OpenAI, we can make some comparisons.

• SWE-bench verified: Google says Gemini 3 scored 76.2 percent, while OpenAI says GPT-5.2 scored 80 percent

• Humanity's Last Exam (HLE), no tools: Google says Gemini 3 scored 37.5 percent; OpenAI says GPT-5.2 scored 34.5 percent

• GPQA Diamond: Google says Gemini 3 scored 91.9 percent; OpenAI says GPT-5.2 scored 92.4 percent.

• AIME 2025, no tools: Google says Gemini scored 95 percent without tools; OpenAI says GPT-5.2 scored 100 percent without tools.

• MMMLU: Google says Gemini 3 scored 91.8 percent; OpenAI says GPT-5.2 scored 89.6 percent.

Overall, it's a mixed bag for both models on benchmark performance — you win some, you lose some. You can check out the full benchmark performance at OpenAI and Google.

GPT-5.2 vs. Gemini 3: Features and availability

Gemini 3 and GPT-5.2 are both series of models, which are then integrated into the Gemini and ChatGPT platforms. And these platforms have different capabilities. For instance, Gemini lets users create both AI-generated photos and videos. ChatGPT lets users create images, but users need to use the Sora app for AI videos. Professional users can also access the new models via Google or OpenAI's API and enterprise systems.

Google does have the edge in one respect: Gemini 3 is integrated into more products, giving users more ways to access the new reasoning model. In addition to the Gemini and Google apps, Gemini 3 powers Google AI Mode. Google has also integrated Gemini 3 into various Google AI Studio projects and tools like NotebookLM.

Ultimately, you can do more with Gemini 3 than you can with GPT-5.2, even if it turns out GPT-5.2 performs better.

GPT-5.2 vs. Gemini 3: Pricing

Price-wise, both AI models are pretty even. For $20 per month, users get expanded access to the GPT-5.2 models within OpenAI's ChatGPT Plus plan (the Pro plan costs $200/month). Google also charges $20 per month for its Google AI Pro plan (and $249.99/month for Google AI Ultra, which comes with Google cloud storage). The two are also pretty evenly priced when it comes to API access:

• GPT‑5.2 is priced at $1.75 per 1 million input tokens and $14 per 1 million output tokens

• Gemini 3 is priced at $2 per 1 million input tokens and $12 per 1 million output tokens

Our recommendation for AI users? Try the two AI models out yourself and see which one best fits your needs and workflow.

Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.

SAVE $200.02: As of Dec. 11, get the 65-inch TCL 4K QLED TV for $499.97 at Amazon, down from its usual price of $699.99. That's a discount of 29% and the lowest price we've seen.

Opens in a new window Credit: Amazon 65-inch TCL 4K QLED TV $499.97 at Amazon
$699.99 Save $200.02   Get Deal

It might be hard to believe, but TVs are relatively cheap these days. If you're trying to find a great one for entertaining this year during the holidays, you can snap one up for a surprisingly low price, with quality and feature ranges that you might not expect for what you pay. We found a great pick that works as either a gift or your new home centerpiece, and it's up for grabs right now at Amazon for the lowest price we've seen.

As of Dec. 11, get the 65-inch TCL 4K QLED TV for $499.97 at Amazon, down from its usual price of $699.99. That's $200.02 off and a discount of 29%. It's also the lowest price we've seen.

SEE ALSO: Best TVs under $500: Low-budget, high-quality TVs in every size

This 4K ultra HD QLED TV is a great pick for detail, with four times the realism of what you can expect from normal HD TVs. It uses QLED tech to enhance its color range, using nearly the entire DCI-P3 color space to give you the richest picture you can get.

It features a 144Hz panel refresh rate for a better, smoother viewing experience no matter whether you're playing games, watching sports, or checking out fast-paced content. Thanks to its TCL AIPQ Pro processor, it can also automatically optimize color, contrast, and clarity on its own to give you a great picture.

With Dolby Atmos spatial audio and voice control support, you'll get fantastic sound as well as be able to navigate the TV with your voice. You'll appreciate that for the built-in smart TV features that let you go without a streaming device as well.

If you need an affordable TV with great quality image and sound, this is a pick you should keep your eye on, especially at this entry price.

The order would create one federal regulatory framework for artificial intelligence, Mr. Trump told reporters in the Oval Office.

Gaming’s biggest night is here. The Game Awards 2025 airs tonight, Dec. 11, at 5 p.m. PT. The show will give out awards to the most notable games of 2025, including the coveted Game of the Year award.

Alongside these awards, we’re also expecting some brand-new game announcements and reveals. Some studios have already teased that they’ll be debuting some surprises tonight, but who knows what the night has in store.

Check out the full list of categories and nominations, and keep reading for the latest news and updates from The Game Awards 2025.

What we're expecting from the big show: Tomb Raider, Resident Evil, and more

Crystal Dynamics teased that it’ll showcase the future of the Tomb Raider franchise. Presumably, it’s going to be the next entry in the franchise.

Capcom is expected to show more footage for Resident Evil Requiem. Fans have speculated that Leon Kennedy would make a return to the franchise, and when the cover was leaked by the PlayStation Store earlier this week, it showed Leon. All we need now is for Capcom to confirm the news during the show.

Sega is teasing something as well, saying that we won’t want to miss it. The legacy gaming company has a few projects in the works, including Virtua Fighter 6 and Stranger Than Heaven, so it’s possible we’ll get a release date for these upcoming titles. There might be a Total War announcement too. On the Atlus side, it’s highly unlikely that Persona 6 will get announced, but we could get more information on the Persona 4 remake.

What we hope to see at the show: Unconfirmed sequels and remakes

Lastly, there’s a strange statue that’s been popping up, and host Geoff Keighley has been teasing something big. Internet sleuths have already dug up that it’s likely Baldur’s Gate 3 dev Larian Games’ next project, which is seemingly a Divinity project.

There have been a few trademark leaks too, which hint at what could be revealed.

An Assassin’s Creed IV Black Flag remake was reported to be in development, and a trademark for Assassin’s Creed Black Flag Resynced popped up a few days ago.

Remedy is making more games in the Control universe, and a recent trademark for Control: Resonant appeared. This could be the sequel to the 2019 game.

We’ll be updating this list during The Game Awards tonight, so be sure to check back for the latest live updates and news!

The House of Mouse recently announced a surprise $1 billion deal with OpenAI, which will bring about 200 Disney characters, props, and other copyrighted items to ChatGPT and Sora, OpenAI's AI chatbot and video app.

So, if you've ever wanted to make a video of yourself hanging out with Deadpool and wielding a lightsaber, you'll be able to do exactly that starting in 2026.

And though many artists and creative professionals have a deep-seated, at-times existential hatred for all things artificial intelligence, Disney CEO Bob Iger said creators don't need to worry. During a CNBC appearance on Thursday, Iger said that "this does not in any way represent a threat to the creators at all."

Well, OK then!

Incredibly, even though OpenAI CEO Sam Altman told CNBC the demand for Disney characters among ChatGPT users is "off the charts," according to the terms of the deal, Disney will pay OpenAI, and not the other way around, to bring Disney intellectual property into ChatGPT-world. Disney will be making a $1 billion equity investment in the AI company, OpenAI revealed in a blog post. As part of the deal, Disney will also become a "major customer" of OpenAI, giving Disney employees access to new AI tools.

Why are creatives threatened by generative AI?

Despite Iger's rosy assessment, the Disney-OpenAI alliance will be seen as terrible news by AI-skeptical artists, who fear that movie studios will replace them with AI tools at the first opportunity. That's why even the smallest hint of AI in movies triggers an immediate backlash from many creative professionals. Even the use of AI to produce movie posters will spark a boycott among artists, many of whom see all generative AI output as slop.

Although, as I've written before, it's not clear that audiences care as much about the use of AI in creative work. Ultimately, the Disney-OpenAI deal is just the latest proof that generative AI in movies, TV, and advertising is a train that can't be stopped. Artists can and will fight the adoption of AI in Hollywood, but the powers that be aren't asking for permission.

In his CNBC appearance with Altman, Iger spoke about the inevitability of AI adoption.

"If it’s going to happen regardless, then we’d rather participate in the rather dramatic growth, rather than just watching it happen and essentially being disrupted by it," Iger said.

Notably, the new agreement does not include the voices of any Disney characters. Including the voice or likeness of the actors behind the characters could put Disney in violation of agreements with writers, artists, and actors, whose unions have been fighting for protections against AI.

As part of the deal, OpenAI and Disney said there's a "licensing fee" for using Disney's characters. But since Disney is the one forking over $1 billion, it seems extremely unlikely that any creatives will see the benefits of this deal. Disney may yet set a precedent that AI companies should pay to license creatives' work, but this isn't that.

"We are not including name and likeness, nor are we including character voices," Iger said on CNBC show Squak on the Street. "And so, in reality, this does not in any way represent a threat to the creators at all, in fact, the opposite. I think it honors them and respects them, in part because there’s a license fee associated with it."

For the record, I am extremely skeptical that "license fee" will trickle down to the artists behind the characters that appear in Sora. I'm equally skeptical that artists, voice actors, and artists will agree that this deal "honors" and "respects them."

By aligning with OpenAI, Disney opens a new front in the AI copyright battle

In Mashable's testing of AI image generators, we've found every major AI image maker will readily produce a deepfake featuring a popular Disney character. It's part of our standard testing process for AI image generators.

So, unsurprisingly, a legal battle is playing out in courts around the world over AI copyright law, with Hollywood heavyweights, authors, and artists pitted against the big AI companies. Some of these cases focus on AI companies using copyrighted works for training without payment or permission. Others focus on the output of those AI tools, many of which readily create deepfakes of popular characters from Star Wars, Pixar, Marvel, and other Disney franchises.

So far, early AI copyright cases have ruled in favor of AI companies. It's been a true David vs. Goliath story so far, with relatively unknown authors going up against tech giants like OpenAI, Google, and Meta.

That's why many artists were delighted when, earlier this year, Disney sued AI image generator Midjourney, calling it "a bottomless pit of plagiarism." The hope was that The House of Mouse's famously formidable legal department might succeed where early class action suits had failed. In the legal battle between artists and AI companies, artists are generally outgunned at every turn. Case in point: In the Kadrey v. Meta class action case, the judge noted in his ruling that the authors' legal team simply made weak arguments, even though the facts of the case seemed so damning. Meta literally pirated the authors' books, and internal messages showed its own employees questioned the legal and ethical implications of their work.

But artists who were hoping Disney might come to the rescue in the AI copyright battle may want to temper their expectations. Disney has now made the bold decision to pay OpenAI for the privilege of turning Disney characters into slop.

It's easy to see why. On the same day we learned about the Disney-OpenAI deal, OpenAI sent a cease-and-desist letter to Google, demanding that Google's AI tools stop creating images and videos that feature Disney characters. By granting OpenAI and Sora an exclusive AI license for these Disney characters, Disney is protecting its copyrights and setting itself up for future legal battles.

None of this is particularly good news for creators and artists worried about AI, however. It should hardly come as a surprise that Iger's best interests don't necessarily align with those of actors, animators, and voice-over artists.

But, Iger says they don't need to worry, so I'm sure everything will be fine.

Yes, everything will be fine.

Everything will be fine.

Everything will be fine.

Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.

SAVE $130: The TCL NXTPAPER 11 Plus Android tablet is on sale at Amazon for $239.99, down from the list price of $369.99. That's a 35% discount that matches the record-low price.

Opens in a new window Credit: TCL TCL NXTPAPER 11 Plus Android tablet $239.99 at Amazon
$369.99 Save $130   Get Deal

A new tablet for 2026 could be the fresh start we all crave. You have tons of options these days if you're looking for a new tablet. But if you're shopping on a budget or just want an affordable model, there's a nice deal on offer today at Amazon.

As of Dec. 11, the TCL NXTPAPER 11 Plus Android tablet is on sale for $239.99 at Amazon, marked down from the standard price of $369.99. That's a 35% discount that works out a nice $130 savings. It also matches the record-low price.

The versatility of the TCL NXTPAPER 11 Plus is one of the standout features of this tablet. TCL designed this tablet with the brand's VeraView mode which means you can use it in regular mode, ink paper, or color paper mode. The ink paper mode utilizes TCL's NXTPAPER 4.0 technology, which aims to give the display the feel of reading physical paper with less eye-stain that could occur with many tablets. This'll be a highlight for those who love to use e-readers like Kindles.

SEE ALSO: Record-low price alert: We've never seen the AirPods Pro 3 this cheap at Walmart

Considering the least expensive Amazon Kindle Colorsoft is priced at $249.99 and doesn't function as a tablet to access email, the web, etc. the TCL NXTPAPER 11 Plus could be a more versatile alternative.

Mashable Senior Editor Stan Schroeder spent some time with an older TCL NXTPAPER and wrote, "From the get-go, their displays really do look like photos printed on actual paper. They’re matte, and thus better to use in sunlight than the displays on most phones and tablets. They appear perfectly still and perfectly flat, for lack of better words."

This deal also includes a flip case and the TCL T-Pen, which magnetically attaches to the case for added convenience.

While the TCL NXTPAPER 11 Plus is sitting at a record-low price of $239.99, consider the upgrade a gift to start off 2026 on the right foot. You'll have access to an impressive display for reading e-books, taking on creative projects, and keeping up with e-mails.

TL;DR: Make peace with PDFs with this lifetime subscription to PDF Converter and Editor, now just $23.99 (reg. $99.99) with code CONVERT through Jan. 11.

Opens in a new window Credit: Acethinker PDF Converter & Editor: Lifetime License $23.99
$99.99 Save $76   Get Deal

PDFs don’t have to be frustrating. Whether you edit PDFs daily or just every now and then, the right tool for the job makes a big difference. That’s where the PDF Converter and Editor app comes into play, offering everything you need to tackle PDFs in one convenient place.

Right now, a lifetime subscription to PDF Converter and Editor can be yours for just $23.99 (reg. $99.99) with code CONVERT through Jan. 11.

Mashable Deals Be the first to know! Get editor selected deals texted right to your phone! Get editor selected deals texted right to your phone! Loading... Sign Me Up By signing up, you agree to receive recurring automated SMS marketing messages from Mashable Deals at the number provided. Msg and data rates may apply. Up to 2 messages/day. Reply STOP to opt out, HELP for help. Consent is not a condition of purchase. See our Privacy Policy and Terms of Use. Thanks for signing up!

No matter how much we try, we can’t escape PDFs. If you’re ready to be prepared every time you encounter one of these frustrating file formats, PDF Converter and Editor can help. This handy tool makes it easy to work with these files and convert them to another format in just a few clicks.

Whether you need to submit a signature or submit a bunch of forms, PDF Converter and Editor helps you work within PDF files and convert them into other file types — including Microsoft Word, Excel, PowerPoint, text, HTML, PNG, or JPG — with just a few taps.

If you need to edit an existing file, PDF Converter and Editor allows you to merge, split, rotate, or delete pages with just a few taps. You can also use OCR to extract and edit text, or even add notes, stamps, text, or a customizable watermark if needed.

Need to submit something as a PDF? That’s just as straightforward with this app, which lets you create a PDF from a Word, Excel, PowerPoint, or image file as well. Converting files won’t sacrifice quality, as original layouts, images, and hyperlinks will all be preserved in the process.

Conquer PDFs for good with this lifetime subscription to PDF Converter and Editor, now just $23.99 (reg. $99.99) with code CONVERT until Jan. 11.

StackSocial prices subject to change.

TL;DR: Learn new concepts in 15 minutes with this lifetime subscription to Headway Premium, on sale now for just $39.99, the best price on the web, with code SUMMARY20 through Jan. 31.

Opens in a new window Credit: Headway Headway Premium: Lifetime Subscription $39.99
$299.95 Save $259.96   Get Deal

Do you have some lofty goals for 2026? Whether you have a lengthy TBR list or you are just hoping to learn some new things next year, Headway Premium allows you to access 15-minute summaries of some of the world’s best nonfiction books in your spare time.

Right now, you can score a lifetime subscription to Headway Premium for just $39.99 with code SUMMARY20 until Jan. 31.

Mashable Deals Be the first to know! Get editor selected deals texted right to your phone! Get editor selected deals texted right to your phone! Loading... Sign Me Up By signing up, you agree to receive recurring automated SMS marketing messages from Mashable Deals at the number provided. Msg and data rates may apply. Up to 2 messages/day. Reply STOP to opt out, HELP for help. Consent is not a condition of purchase. See our Privacy Policy and Terms of Use. Thanks for signing up!

If you’ve got 15 minutes, Headway Premium can teach you something new. Ditch your doomscrolling and start using this app to read or listen to 15-minute summaries of bestselling nonfiction books.

More than 15 million people are already learning with Headway, picking up key ideas and insights from books in their spare time. Choose between listening to professionally narrated audio summaries or reading written ones, depending on your preference.

There are over 2,000 summaries already available, with new ones added each month. Choose from topics such as personal development and business strategies, or health and wellness. Headway’s gamified learning makes the process even more fun, as you can earn achievements, track your progress, and master new skills while you work on your self-growth.

Heading on a trip? You can enjoy summaries offline as well. While it doesn’t replace reading the whole book from cover to cover, it is a great way to discover new subjects you enjoy.

Learn new things in 2026 and beyond with this lifetime subscription to Headway Premium, now only $39.99, the best price on the web, with code SUMMARY20 through Jan. 31.

StackSocial prices subject to change.

Do Kwon, who designed the virtual currencies Luna and TerraUSD, which plunged in 2022, had pleaded guilty to fraud.

President Donald Trump announced this week that he intends to sign an executive order designed to stop states from passing laws to regulate artificial intelligence.

A new public service announcement, however, challenges Trump's position by attempting to illustrate how children and teens have already been harmed by AI chatbots, in the absence of robust state and federal regulation.

SEE ALSO: 5 strategies to avoid ChatGPT dependency

The spot was commissioned by the child safety advocacy groups Heat Initiative, ParentsTogether Action, and Design It For Us, and was narrated by actress Juliette Lewis.

Creepy-looking humans are cast as the voices and faces of real AI chatbots that have allegedly shared dangerous information with young users who engaged with them, like how to harm themselves and hide an eating disorder from their parents. Three examples reference instances in which ChatGPT allegedly coached young users into attempting suicide. Experts recently reviewed major AI chatbots and concluded they're not safe for teens to use for mental health discussions.

OpenAI, the maker of ChatGPT, is being sued by multiple families of teens who died by suicide after heavy engagement with the chatbot. The company recently denied responsibility for the death of Adam Raine, a 16-year-old who talked to ChatGPT about his suicidal feelings and killed himself earlier this year.

"As parents, we do everything in our power to protect our children from harm, but how do we protect them from powerful technologies designed to exploit their vulnerabilities for profit?" said Megan Garcia, whose son, Sewell Setzer III, died by suicide in the wake of developing an intense relationship with a chatbot on Character.AI. The company has since shut down teen chats on the platform.

"If state AI regulations are blocked and AI companies are allowed to keep building untested and dangerous products, I'm afraid that many more families will endure the agony of losing a child. We cannot accept losing one more child to AI harms," Garcia said.

If you're feeling suicidal or experiencing a mental health crisis, please talk to somebody. You can call or text the 988 Suicide & Crisis Lifeline at 988, or chat at 988lifeline.org. You can reach the Trans Lifeline by calling 877-565-8860 or the Trevor Project at 866-488-7386. Text "START" to Crisis Text Line at 741-741. Contact the NAMI HelpLine at 1-800-950-NAMI, Monday through Friday from 10:00 a.m. – 10:00 p.m. ET, or email info@nami.org. If you don't like the phone, consider using the 988 Suicide and Crisis Lifeline Chat. Here is a list of international resources.

Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.

Today, Disney and OpenAI announced a surprising partnership, which will allow ChatGPT users to generate content with Disney's iconic characters.

On the heels of that news, Variety reported that Disney sent Google a cease-and-desist notice on Wednesday evening, threatening legal action over the search giant's alleged use of Disney's intellection property in its AI models. Disney is accusing Google of copyright infringement on a “massive scale,” claiming Google's AI models “commercially exploit and distribute” infringing media.

Disney has reportedly accused Google of infringing on Disney properties such as Star Wars, Frozen, The Lion King, Moana, The Little Mermaid, Deadpool, and The Guardians of the Galaxy. Disney has also accused Google of encouraging users to take part in Generative AI trends such as prompts that create images of "action figures" depicting Disney-owned characters. According to Variety, images of figurines depicting Darth Vader, Deadpool, Homer Simpson, and Elsa from Frozen were included in examples in the cease-and-desist letter.

Disney's recent actions are certainly interesting. Just this past June, Disney filed a lawsuit against AI image generation company Midjourney over the use of Disney's IP in AI-generated content. At the time, Disney lawyers called Midjourney "a bottomless pit of plagiarism."

So, why has Disney partnered with one, and only one, AI company? By granting OpenAI an exclusive license to use its characters, OpenAI will now be able to argue that Google and other AI companies are violating that license.

When OpenAI launched its AI video platform Sora in October, the company was immediately criticized for allowing users to generate video content utilizing other brands' copyrighted characters. OpenAI eventually announced that any company could opt out of Sora inclusion on a character-by-character basis.

And in Mashable's testing of AI image and video generators, we found that most of these AI tools will readily produce deepfakes featuring Star Wars and Marvel characters.

Now, Disney has demanded that Google “immediately cease further copying, publicly displaying, publicly performing, distributing, and creating derivative works of Disney’s copyrighted characters” in “outputs of Google’s AI Services, including through YouTube’s mobile app, YouTube Shorts and YouTube.”

In addition, Disney wants Google to put safeguards in place "to ensure that no future outputs infringe Disney works."

In effect, Disney is putting the AI world on notice.

Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.

Dropped onto HBO Max in late Nov. with almost zero promotion, the gay hockey series Heated Rivalry is everywhere these days. The show's handsome stars, Hudson Williams and Connor Storrie, are dominating social feeds, with a growing legion of fans salivating over the chemistry between their characters, two extremely talented and closeted NHL players. As of Dec. 11, Heated Rivalry sits as the third most-watched series on HBO Max, behind only Mad Men and IT: Welcome to Derry.

What is it about this Canadian show that has captured the public's attention in a way few other recent series have? Yes, it's a rare gay love story on TV, and it depicts the often-untold tales of LGBTQ athletes. The acting is good, as are the production values. Undeniably, though, it's the show's depictions of sex, mostly between Williams' and Storrie's characters, that's pulling in viewers. Their trysts feel different than same-sex "lovemaking" in other notable LGBTQ shows, like Queer as Folk, The L Word, and, more recently, Heartstopper.

The nature of the characters' relationship is certainly part of it — their sexual congress is beyond illicit. If the public were to find out the League's two best players were sleeping together (and were, maybe, in love?), it could wreck their careers and even endanger their lives. The high stakes equate to intense passion.

SEE ALSO: 9 LGBTQ creators discuss not backing down from Pride

It's also to the testament of series creator Jacob Tierney that the sex comes off as authentic and hot. He cast two charismatic leads (who seem to actually like each other in real life) and filmed them beautifully in luxe hotel rooms. Heated Rivalry's lighting is warm and sensual, though HBO Max did darken the show (inadvertently, Tierney maintains), including its sex scenes, from the original Canadian version. The camera stays on the men during the actual intercourse, with no burning candles or hand-holding for suggestion. When the characters of Shane and Ilya give in to their passion, it's less Red, White, & Royal Blue (e.g., soap opera-style timidity) and much more Blue Is the Warmest Color (close-ups, noises, and occasional awkwardness).

This Tweet is currently unavailable. It might be loading or has been removed.

Zachary Zane, Grindr's sex and relationship expert, heaped praise on Heated Rivalry's depiction of gay lust:

"First off, the show masterfully builds suspense and sexual tension, perhaps more so than previous depictions of queer sex on screen. It adopts a slow-burn approach, characterized by constant furtive glances and intense pining between Ilya and Shane, with Shane's longing being particularly noticeable. The viewer can see the sexual tension and desire on their pretty faces.

Two, the production value is also incredibly high. The soft lighting and shadows add a sensual and suspenseful element we didn't get nearly as much of in past queer shows.

Three, the dynamic between Ilya and Shane is a major draw. Their enemies-to-lovers trope is an inherently popular and arousing romance staple, but it's more than that. Shane leans a bit submissive, being coy about his wants while dancing around his anxieties. Ilya is this powerhouse, emotionally-avoidant dom with a sexy accent, deep voice, and adventurous mindset for public trysts. If only we could see their Grindr tags.

While sexual fantasies and desires naturally vary among queer men, this particular dom/sub dynamic is a common interest amongst most gay and bisexual men, who can identify with either Ilya, Shane, or perhaps even both."

Heated Rivalry's fourth episode (of six) drops tomorrow, Dec. 12, on HBO Max.

Today, OpenAI launched GPT-5.2, touting its stronger safety performance in regard to mental health.

"With this release, we continued our work to strengthen our models' responses in sensitive conversations⁠, with meaningful improvements in how they respond to prompts indicating signs of suicide or self-harm, mental health distress, or emotional reliance on the model," OpenAI's blog post states.

OpenAI has recently been hit with criticism and lawsuits, which accuse ChatGPT of contributing to some users' psychosis, paranoia, and delusions. Some of those users died by suicide after lengthy conversations with the AI chatbot, which has had a well-documented problem with sycophancy.

In response to a wrongful death lawsuit concerning the suicide of 16-year-old Adam Raine, OpenAI denied that the LLM was responsible, claimed ChatGPT directed the teenager to seek help for his suicidal thoughts, and stated that the teenager "misused" the platform. At the same time, OpenAI pledged to improve how ChatGPT responds when users display warning signs of self-harm and mental health crises. As many users develop emotional attachments to AI chatbots like ChatGPT, AI companies are facing growing scrutiny for the safeguards they have in place to protect users.

Now, OpenAI claims that its latest ChatGPT models will offer "fewer undesirable responses" in sensitive situations.

In the blog post announcing GPT-5.2, OpenAI states that GPT-5.2 scores higher on safety tests related to mental health, emotional reliance, and self-harm compared to GPT-5.1 models. Previously, OpenAI has said it's using "safe completion," a new safety-training approach that balances helpfulness and safety. More information on the new models' performance can be found in the 5.2 system card.

Credit: Screenshot: OpenAI

However, the company has also observed that GPT-5.2 refuses fewer requests for mature content, especially sexualized text. But this apparently doesn't impact users OpenAI knows to be underage, as the company states that its age safeguards "appear to be working well." OpenAI applies additional content protections for minors, including reducing access to content containing violence, gore, viral challenges, roleplay of sexual, romantic, or violent nature, and "extreme beauty standards."

An age prediction model is also in the works, which will allow ChatGPT to estimate its users' ages to help provide more age-appropriate content for younger users.

Earlier this fall, OpenAI introduced parental controls in ChatGPT, including monitoring and restricting certain types of use.

OpenAI isn't the only AI company accused of exacerbating mental health issues. Last year, a mother sued Character.AI after her son's death by suicide, and another lawsuit claims children were severely harmed by that platform's "characters." Character.AI has been declared unsafe for teens by online safety experts. Likewise, AI chatbots from a variety of platforms, including OpenAI, have been declared unsafe for teens' mental health according to child safety and mental health experts.

If you're feeling suicidal or experiencing a mental health crisis, please talk to somebody. You can call or text the 988 Suicide & Crisis Lifeline at 988, or chat at 988lifeline.org. You can reach the Trans Lifeline by calling 877-565-8860 or the Trevor Project at 866-488-7386. Text "START" to Crisis Text Line at 741-741. Contact the NAMI HelpLine at 1-800-950-NAMI, Monday through Friday from 10:00 a.m. – 10:00 p.m. ET, or email info@nami.org. If you don't like the phone, consider using the 988 Suicide and Crisis Lifeline Chat. Here is a list of international resources.

Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.

A new technology release from OpenAI is supposed to top what Google recently produced. It also shows OpenAI is engaged in a new and more difficult competition.

OpenAI has announced its launching GPT-5.2 effective immediately, with access rolling out in phases starting Dec. 11.

This is the second major update to GPT-5, the long-awaited OpenAI model that launched on Aug. 7.

In a blog post announcing the new series of ChatGPT models, the company called GPT-5.2 the "most capable model series yet for professional knowledge work."

This Tweet is currently unavailable. It might be loading or has been removed.

GPT-5.2 is actually a series of models; the series itself is part of the larger GPT-5 series. Here's what we're looking at:

• GPT-5.2 Instant

• GPT-5.2 Thinking

• GPT-5.2 Pro

When can you start using GPT-5.2?

With a Dec. 11 release date, GPT-5.2 is available now — though not all users will see it listed as an available model in ChatGPT.

As with previous model rollouts, OpenAI will first deploy the model to paid users on Plus, Pro, Go, Business, and Enterprise plans.

Even for paying customers, OpenAI tends to make its new models available in phases — so if it's not available immediately via the OpenAI API or ChatGPT, keep checking.

"We deploy GPT‑5.2 gradually to keep ChatGPT as smooth and reliable as we can," the company wrote. "If you don’t see it at first, please try again later. In ChatGPT, GPT‑5.1 will still be available to paid users for three months under legacy models, after which we will sunset GPT‑5.1."

If you have a paid ChatGPT account, go to ChatGPT.com to see if the model is available. Eventually, all ChatGPT users should have free access to GPT-5.2, though that access will be limited on free membership plans.

GPT-5.2 benchmarks and performance

OpenAI published a new system card that shows GPT-5.2 makes iterative improvements on key benchmarks compared to GPT-5 and 5.1. In addition, OpenAI said that GPT-5.2 hallucinates less than previous models.

• GDPval: GPT-5.2 scored 70.9 percent, while GPT-5 scored 38.8 percent

• SWE-Bench Pro: GPT-5.2 scored 55.6 percent, while GPT-5 scored 50.8 percent

• AIME 2025 math: GPT-5.2 scored 100 percent, while GPT-5 scored 94 percent

• FrontierMath (Tier 1–3): GPT-5.2 scored 40.3 percent, while GPT-5 scored 31.0 percent

Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.

The deal is a watershed for Hollywood, which has been trying to sort through the possible harms and upsides of generative artificial intelligence.

OpenAI announced today that it's launching GPT-5.2, the newest model in its GPT-5 series. The new model will start rolling out immediately, with paid ChatGPT customers getting access first.

In a blog post announcing the new model — which is actually a series of models, comprised of GPT‑5.2 Instant, GPT-5.2 Thinking, and GPT-5.2 Pro — OpenAI said that GPT-5.2 makes noticeable improvements in math and science, imaging, coding, handling agentic tasks, and overall accuracy. The company called GPT-5.2 its "most capable model series yet for professional knowledge work."

The new model comes at a difficult time for OpenAI, which is rumored to be in a "code red" state over stronger competition from rivals like Google Gemini and spreading fears of an AI bubble.

Ever since it launched ChatGPT in 2022, OpenAI has been securely on top of the AI industry. However, the company is in an increasingly precarious position. Google has an almost unfathomable amount of training data at its disposal, and Google AI products like Gemini 3, Veo 3, and Nano Banana have outperformed GPT-5, the new model OpenAI launched earlier this year, in many respects.

Still, ChatGPT is by far the most popular AI chatbot in the world, with an estimated 700 million weekly active users.

How to try GPT-5.2

The new GPT-5.2 models will start rolling out immediately, though access may not be available right away to all users. As per usual, OpenAI will launch the models to paid users on the Plus, Pro, Go, Business, and Enterprise accounts.

As of this writing, GPT-5.2 was not yet available for this reporter, and the rollout will likely happen in phases.

"We deploy GPT‑5.2 gradually to keep ChatGPT as smooth and reliable as we can; if you don’t see it at first, please try again later," OpenAI wrote in a blog post. "In ChatGPT, GPT‑5.1 will still be available to paid users for three months under legacy models, after which we will sunset GPT‑5.1."

OpenAI says GPT-5.2 makes key improvements in safety, accuracy, and performance benchmarks

The AI industry relies on standardized benchmark tests to demonstrate how well models perform, and companies like OpenAI also have their own internal tests. In addition, AI leaderboards like LMArena let users compare and rank various AI models. While GPT-5.2 has already appeared near the top of LMArena's AI coding leaderboard, it will take more time to see how users rate the new series of models against the competition. However, OpenAI released a new model card for GPT-5.2 on Dec. 11, which shows that the model makes across-the-board improvements in a variety of areas, which isn't surprising.

Most notably, OpenAI says that GPT-5.2 is more accurate and will produce fewer hallucinations compared to GPT-5.1. OpenAI's documentation states that GPT-5.2 Thinking has an average hallucination rate of 10.9 percent, compared to 16.8 percent and 12.7 percent for GPT-5 Thinking and GPT-5.1 Thinking, respectively. When GPT-5.2 is given access to the web via a browser, its hallucination rate drops to 5.8 percent.

In its blog post, OpenAI also states that GPT-5.2 scores more highly on benchmark tests for coding, science and math, performing economically valuable tasks, computer vision, and agentic work involving third-party tools. OpenAI also highlighted GPT-5.2's improved abilities with spreadsheets, in particular.

OpenAI says GPT-5.2 is safer for users with mental health problems

Lately, OpenAI has been accused of endangering ChatGPT users with mental health issues. Due to well-documented sycophancy problems, ChatGPT reportedly encouraged delusions and conspiratorial thinking on some users, who later died by suicide. OpenAI is now facing wrongful death suits, including a new suit that was just revealed for the first time today by the Wall Street Journal, in which a ChatGPT user killed himself shortly after killing his own mother.

OpenAI says that according to its internal tests, GPT-5.2 has a better response to users with mental health problems.

"With this release, we continued our work to strengthen our models’ responses in sensitive conversations⁠, with meaningful improvements in how they respond to prompts indicating signs of suicide or self harm, mental health distress, or emotional reliance on the model. These targeted interventions have resulted in fewer undesirable responses in both GPT‑5.2 Instant and GPT‑5.2 Thinking as compared to GPT‑5.1 and GPT‑5 Instant and Thinking models."

Mashable has not been able to independently verify these results, and the GPT-5.2 system card has scant details on how safety performance was measured in this context.

For more information, check out the OpenAI blog post announcing GPT-5.2 or read the new GPT-5.2 system card.

If you're feeling suicidal or experiencing a mental health crisis, please talk to somebody. You can call or text the 988 Suicide & Crisis Lifeline at 988, or chat at 988lifeline.org. You can reach the Trans Lifeline by calling 877-565-8860 or the Trevor Project at 866-488-7386. Text "START" to Crisis Text Line at 741-741. Contact the NAMI HelpLine at 1-800-950-NAMI, Monday through Friday from 10:00 a.m. – 10:00 p.m. ET, or email info@nami.org. If you don't like the phone, consider using the 988 Suicide and Crisis Lifeline Chat. Here is a list of international resources.

Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.

SAVE $1,000: As of Dec. 11, the Bluetti Elite 200 V2 solar generator with two 200W solar panels is just $1,299 at Amazon. That's 43% or a full $1,000 off its list price.

Opens in a new window Credit: Bluetti Bluetti Elite 200 V2 solar generator + two 200W solar panels $1,299 at Amazon
$2,299 Save $1,000   Get Deal

Solar generators (aka portable power stations) can be a real lifesaver in many circumstances. Maybe you have a camping trip coming up, but you still need to access your computer for work. Or maybe you live in an area prone to power outages. In any case, having immediate access to one of these gadgets can keep you up and running. They're not the cheapest investments, so when you find a deal — like this Bluetti Elite 200 V2 bundle — it's worth grabbing.

As of Dec. 11, you can pick up the Bluetti Elite 200 V2 solar generator along with two 200W solar panels for only $1,299 at Amazon. That's 43% or $1,000 in savings. That's also cheaper than buying just the solar generator on its own at full price (reg. $1,699).

The Elite 200 V2 packs a 2,073Wh-capacity battery with a 2,600W (3,900W Power Lifting) AC output. That's enough to power nine devices simultaneously via four AC outlets, two 15W USB-A ports, two 100W USB-C ports, and a 120W car port. Meanwhile, it weighs just over 50 pounds and is the size of a small cooler. It's hefty, but still small enough to fit snugly in a corner of your home or portable enough to take on camping trips, tailgates, and other off-grid adventures.

The power station itself can be charged up via a typical AC outlet, with a car port, or with solar power. This bundle includes two 200W solar panels, so you'll be all set. With a 17-year lifespan, it's a smart long-term investment for your home and adventures. And knowing it will last so long definitely makes the price more justified. Save $1,000 while this deal is live.

Governments are studying the decision to prohibit youths from using platforms like Facebook and TikTok as worries grow about the potential harm they cause.

Android users can now share live video of their environment while reaching out for emergency services, providing more context and assistance in potentially dire situations.

Launched on Wednesday, Emergency Live Video on Android lets emergency dispatchers request a live video feed while talking to a person who may be in danger. When initiated, callers are prompted with a single-tap button on their screen, which then provides a live video feed to dispatchers who can assess the environment or provide more specific medical assistance, like CPR instructions.

SEE ALSO: People negatively judge others who glitch on video calls, according to study

A Google blog post states that the resource works for both emergency calls and texts, and the footage is encrypted by default. The company also says it is looking to partner with global public safety organizations to expand the feature's reach.

Google has launched a variety of emergency safety features across its Android devices, including crash and fall detection, Emergency SOS, and satellite-based services. With the most recent Android 16, Google also added new accessibility features, parental controls, and the ability to mark communications as urgent.

Emergency Live Video has already rolled out to U.S. devices and emergency service calls in select regions of Germany and Mexico. In order to use the feature, devices must be running Android 8+ with Google Play services.