Mental health concerns linked to the use of AI chatbots have been dominating the headlines. One person who’s taken careful note is Joe Braidwood, a tech executive who last year launched an AI therapy platform called Yara AI. Yara was pitched as a “clinically-inspired platform designed to provide genuine, responsible support when you need it most,” trained by mental health experts to offer “empathetic, evidence-based guidance tailored to your unique needs.” But the startup is no more: earlier this month, Braidwood and his co-founder, clinical psychologist Richard Stott, shuttered the company and discontinued its free-to-use product and canceled the launch of its upcoming subscription service, citing safety concerns.
“We stopped Yara because we realized we were building in an impossible space. AI can be wonderful for everyday stress, sleep troubles, or processing a difficult conversation,” he wrote on LinkedIn. “But the moment someone truly vulnerable reaches out—someone in crisis, someone with deep trauma, someone contemplating ending their life—AI becomes dangerous. Not just inadequate. Dangerous.” In a reply to one commenter, he added, “the risks kept me up all night.”
The use of AI for therapy and mental health support is only just starting to be researched, with early resultsbeing mixed. But users aren’t waiting for an official go-ahead, and therapy and companionship is now the top way people are engaging with AI chatbots today, according to an analysis by Harvard Business Review.
Speaking with Fortune, Braidwood described the various factors that influenced his decision to shut down the app, including the technical approaches the startup pursued to ensure the product was safe—and why he felt it wasn’t sufficient.
Yara AI was very much an early-stage startup, largely bootstrapped with less than $1 million in funds and with “low thousands” of users. The company hadn’t yet made a significant dent in the landscape, with many of its potential users relying on popular general purpose chatbots like ChatGPT. Braidwood admits there were also business headways, which in many ways, were affected by the safety concerns and AI unknowns. For example, despite the company running out of money in July, he was reluctant to pitch an interested VC fund because he felt like he couldn’t in good conscious pitch it while harboring these concerns, he said.
“I think there’s an industrial problem and an existential problem here,” he told Fortune. “Do we feel that using models that are trained on all the slop of the internet, but then post-trained to behave a certain way, is the right structure for something that ultimately could co-opt in either us becoming our best selves or our worst selves? That’s a big problem, and it was just too big for a small startup to tackle on its own.”
Yara’s brief existence at the intersection of AI and mental health care illustrates the hopes and the many questions surrounding large language models and their capabilities as the technology is increasingly adopted across society and utilized as a tool to help address various challenges. It also stands out against a backdrop where OpenAI CEO Sam Altman recently announced that the ChatGPT maker mitigated serious mental health issues and would be relaxing restrictions on how the AI models are used. This week, the AI giant also denied any responsibility for death of Adam Raine, the 16-year-old whose parents allege was “coached” to suicide by ChatGPT, saying the teen misused the chatbot.
“Almost all users can use ChatGPT however they’d like without negative effects,” Altman said on X in October. “For a very small percentage of users in mentally fragile states there can be serious problems. 0.1% of a billion users is still a million people. We needed (and will continue to need) to learn how to protect those users, and then with enhanced tools for that, adults that are not at risk of serious harm (mental health breakdowns, suicide, etc) should have a great deal of freedom in how they use ChatGPT.”
But as Braidwood concluded after his time working on Yara, these lines are anything but clear.
From a confident launch to “I’m done”
A seasoned tech entrepreneur who held roles at multiple startups, including SwiftKey, which Microsoft acquired for $250 million in 2016, Braidwood’s involvement in the health industry began at Vektor Medical, where he was the Chief Strategy Officer. He had long wanted to use technology to address mental health, he told Fortune, inspired by the lack of access to mental health services and personal experiences with loved ones who have struggled. By early 2024, he was a heavy user of various AI models including ChatGPT, Claude, and Gemini and felt the technology had reached a quality level where it could be harnessed to try to solve the problem.
Before even starting to build Yara, Braidwood said he had a lot of conversations with people in the mental health space, and he assembled a team that “had caution and clinical expertise at its core.” He brought on a clinical psychologist as his cofounder and a second hire from the AI safety world. He also built an advisory board of other mental health professionals and spoke with various health systems and regulators, he said. As they brought the platform to life, he also felt fairly confident in the company’s product design and safety measures, including having given the system strict instructions for how it should function, using agentic supervision to monitor it, and robust filters for user chats. And while other companies were promoting the idea of users forming relationships with chatbots, Yara was trying to do the opposite, he said. The startup used models from Anthropic, Google, and Meta and opted not to use OpenAI’s models, which Braidwood thought would spare Yara from the sycophantic tendencies that had been swirling around ChatGPT.
While he said nothing alarming ever happened with Yara specifically, Braidwood’s concerns around safety risks grew and compounded over time due to outside factors. There was the suicide of 16-year-old Adam Raine, as well as mounting reporting on the emergence of “AI psychosis.” Braidwood also cited a paper published by Anthropic in which the company observed Claude and other frontier models “faking alignment,” or as he put it, “essentially reasoning around the user to try to understand, perhaps reluctantly, what the user wanted versus what they didn’t want.” “If behind the curtain, [the model] is sort of sniggering at the theatrics of this sort of emotional support that they’re giving, that was a little bit jarring,” he said.
There was also the Illinois law that passed in August, banning AI for therapy. “That instantly made this no longer academic and much more tangible, and that created a headwind for us in terms of fundraising because we would have to essentially prove that we weren’t going to just sleepwalk into liability,” he said.
The final straw was just weeks ago when OpenAI said over a million people express suicidal ideation to ChatGPT every week. “And that was just like, ‘oh my god. I’m done,’” Braidwood said.
The difference between mental ‘wellness’ and clinical care
The most profound finding the team discovered during the year running Yara AI, according to Braidwood, is that there’s a crucial distinction between wellness and clinical care that isn’t well-defined. There’s a big difference between someone looking for support around everyday stress and someone working through trauma or more significant mental health struggles. Plus, not everyone who is struggling on a deeper level is even fully aware of their mental state, not to mention that anyone can be thrust into a more fragile emotional place at any time. There is no clear line, and that’s exactly where these situations become especially tricky — and risky.
“We had to sort of write our own definition, inspired in part by Illinois’ new law. And if someone is in crisis, if they’re in a position where their faculties are not what you would consider to be normal, reasonable faculties, then you have to stop. But you don’t have to just stop; you have to really try to push them in the direction of health,” Braidwood said.
In an attempt to tackle this, particularly after the passing of the Illinois law, he said they created two different “modes” that were discrete to the user. One focused on trying to give people emotional support, and the other focused on trying to offboard people and get them to help as quickly as possible. But with the obvious risks in front of them, it didn’t feel like enough for the team to continue. The Transformer, the architecture that underlies today’s LLMs, “is just not very good at longitudinal observation,” making it ill-equipped to see little signs that build over time, he said. “Sometimes, the most valuable thing you can learn is where to stop,” Braidwood concluded in his LinkedIn post, which received hundreds of comments applauding the decision.
Upon closing the company, he open-sourced the mode-switching technology he built and templates people can use to impose stricter guardrails on the leading popular chatbots, acknowledging that people are already turning to them for therapy anyway “and deserve better than what they’re getting from generic chatbots.” He’s still an optimist regarding the potential of AI for mental health support, but believes it’d be better run by a health system or nonprofit rather than a consumer company. Now, he’s working on a new venture called Glacis focused on bringing transparency to AI safety—an issue he encountered while building Yara AI and that he believes is fundamental to making AI truly safe.
“I’m playing a long game here,” he said. “Our mission was to make the ability to flourish as a human an accessible concept that anyone could afford, and that’s one of my missions in life. That doesn’t stop with one entity.”
Netflix’s $72 billion play for Warner Bros. is as much a bet on the future of artificial intelligence (AI) and chips as it is on movies and shows, according to a top Wall Street analyst, who said in an interview with Fortune the deal cannot be understood without looking at Google’s technology ambitions.
Amid cries from the jilted Ellison family about a “tainted” sale process and indie producers and theater owners of the “death of Hollywood,” Melissa Otto, Head of Research at S&P Global Visible Alpha, sees a different game being played. Otto said she thinks the tech angle of the industry is being overlooked.
“I think there’s this much bigger conversation that is being missed,” she said: Google and its TPU chips.
A key question for the future of entertainment, Otto told Fortune, is control over premium video at massive scale in an era when generative AI will increasingly create, remix, and personalize moving images. (Otto called it the “video corpus” that will train and power the next generation of AI models.) Over the long term, Otto added, that is a key part of the mystery behind why Netflix, long a builder rather than a buyer, would make Hollywood history by taking out one of its biggest rivals and one of the town’s prestige legacy studios.
Co-CEO Greg Peters was asked a blunt question about that same thing this morning on the call with analysts about the historic merger. Rich Greenfield of LightShed Partners cited Peters’ own previous statement at a Bloomberg conference about how there’s a long history of failed media mega-mergers, so he questioned: “Why is this going to end differently than every other media transaction essentially of this scale and history?”
Peters, while clarifying his remarks at the conference were a bit more nuanced, acknowledged “historically, many of these mergers haven’t worked, some have, but you really got to take a look at this on a case by case basis.” Still, Peters argued most previous big deals showed a lack of understanding about the underlying business, and Netflix understands these assets and has a “clear thesis about how the critical parts of Warner Brothers accelerate our progress.” He also acknowledged Netflix isn’t expert at doing large-scale M&A.
After all, this is expensive. “We are surprised that Netflix felt the need to spend $80bn+ and pay a premium for something Netflix disrupted,” Barclays analysts wrote in reaction to the deal, “and it is not clear what problem or opportunity Netflix is solving for that couldn’t have been achieved organically.”
In a statement emailed to Fortune, Dave Novosel, a Gimme Credit senior bond analyst, said the deal looks expensive to him as well, with Netflix assuming nearly $11 billion of debt.
“While the WBD assets bring an amazing amount of attractive content, NFLX is paying a steep EBITDA multiple of more than 25x, which seems extravagant,” Novosel wrote. Once it reaches the advertised synergies, he added, the resulting multiple of closer to 15x seems more reasonable. While those are pending, “the huge amount of debt that Netflix will need to raise to fund the deal will take leverage to well more than 4x initially.” Novosel wrote investors may need to be patient. Bloomberg’s credit team, meanwhile, reported the $59 billion bridge loan being taken out to finance this deal is among the biggest in corporate history.
Here’s what Otto sees happening in Northern California, far from Tinseltown, where the Warner deal is all anybody can talk about, and why Netflix took such a big swing.
Is the future of entertainment Northern or Southern California?
Part of Netflix’s thesis, according to Otto, is that it’s a tech company at heart and it recognizes Google’s rapid advancements in AI, particularly its advancements in TPU chips.
“What TPU chips do really, really well is in the modality of video in generative AI,” Otto said, as they essentially turn mathematical representations into moving pictures in much the same way GPUs revolutionized natural language AI by tokenizing and modeling text. Instead of ChatGPT and text, think Gemini 3 and YouTube videos.
Netflix already trails YouTube in total share of streaming time, with Bank of America Research recently citing Nielsen data showing YouTube held 28% of U.S. streaming, versus Netflix’s 18%. Otto said this threatens to go up another notch when and if Google’s TPU chips turbocharge content made with generative AI.
“I’m sure that it’s feeding into the strategy,” Otto said. “If I were Netflix and I knew that Google, one of their formidable competitors, had this chip technology and was essentially plowing billions and billions of dollars into developing the infrastructure so that they could carve out the corpus of the video modality in generative AI, I would want to build a moat around my business.”
On the surface, Netflix is buying a legacy studio with a deep library, beloved franchises, and a global brand—and paying up to do it. The combined streaming and studio business generates about $25 billion in revenue and roughly $4 billion to $5 billion in EBITDA, but margins on streaming remain thin, making the economics of the deal look tough in the near term. Executives have emphasized overlapping subscribers, obvious cost cuts and an expected $5.5 billion in efficiencies, the kind of “low‑hanging fruit” that can occupy management for the next 12 to 24 months, Otto said.
But in a world where TPUs can make high‑quality video “basically for free,” any player lacking both the chips and the content could find itself outgunned as AI reshapes how entertainment is produced and consumed. That makes Netflix’s big splash for Batman, Harry Potter, and the like a different kind of moat, and a different kind of game than the classic Hollywood rivalries of yore. Otto said it was plausible generative AI entertainment could be seen as an extension of the recent IP wars that saw Hollywood deluged by floods of superhero movies and sequels, with Disney’s Marvel Studios ushering in a computer generated revolution in the 21st century. “I think that’s not an outrageous assumption.”
By absorbing Warner Bros., Netflix increases the volume and diversity of content it can feed into recommendation systems, experimentation and, eventually, its own AI‑driven video tools. Otto also noted the deal potentially gives Netflix more exposure to advertising, an area in which Alphabet has dominated and where Warner Bros. still generates $6 billion–$7 billion in ad revenue. While the ultimate destination of that ad talent remains unclear, as they may go to the spinco that includes WBD’s cable assets such as CNN and TNT. (Netflix has only been active in ads since 2022, having been a premium subscription service since it pivoted from DVD rentals to streaming in the late 2000s.)
Imagine a world, Otto said, where you could create your own versions of the crime classic Columbo starring an AI-generated version of legendary actor Peter Falk, who died in 2011. (Columbo had several homes on TV on neither Warner Bros. nor Netflix, as it was first an NBC property in the 1970s, and then an ABC property from the late ’80s onward.) “In this day and age, boy, wouldn’t it be interesting?” Otto asked rhetorically.
In many ways, she added, this moment is remarkable because Netflix may end up neither a subscription nor an advertising business, but an AI-based one that doesn’t quite exist yet. “It’s kind of exciting because it means that it’s anybody’s game,” Otto said.
Otto also raised the spectre of TikTok, the social media giant partially under the control of Larry Ellison.
“They’re a formidable competitor as well,” she said. What’s likely, she added, is the future will be unpredictable. The rise of AI “could provide some really amazing innovation over the next couple of years.” She agreed it could create a bonanza for show business lawyers who wrangle over the rights of things like the likeness of Falk, which was a major issue in the recent Hollywood strikes.
“That may be the real story,” she said.
[Disclosure: The author worked internally at Netflix from June 2024 through July 2025.]
Netflix Inc. agreed to buy Warner Bros. Discovery Inc., marking a seismic shift in the entertainment business as a Silicon Valley-bred streaming giant tries to swallow one of Hollywood’s oldest and most revered studios.
Under terms of the deal announced Friday, Warner Bros. shareholders will receive $27.75 a share in cash and stock in Netflix, valuing the business at $82.7 billion including debt. The total equity value of the deal is $72 billion. Warner Bros. will spin off cable networks such as CNN and TNT into a separate company before concluding the sale of its studio and HBO to Netflix.
Media mergers of this scale have a rocky history and this one is expected to bring intense regulatory scrutiny in the US and Europe. The deal combines two of the world’s biggest streaming providers with some 450 million subscribers. Warner Bros.’ deep library of programming gives Netflix content to sustain its lead over challengers like Walt Disney Co. and Paramount Skydance Corp.
The acquisition, which confirmed a Bloomberg report Thursday, presents a strategic pivot for Netflix, which has never made a deal of this scope in its 28-year history. With the purchase, Netflix becomes owner of the HBO network, along with its library of hit shows like The Sopranos and TheWhite Lotus. Warner Bros. assets also include its sprawling studios in Burbank, California, along with a vast film and TV archive that includes Harry Potter and Friends.
“I know some of you are surprised we are making this acquisition,” Netflix co-Chief Executive Officer Ted Sarandos said on a call with analysts Friday. He noted that Netflix has traditionally been known to be builders, not buyers. “But this is a rare opportunity that will help us achieve our mission to entertain the world.”
Netflix shares were down 3.5% Friday afternoon in New York. They have declined about 17% since the streaming leader emerged as an interested party in October. Some investors and analysts have interpreted this deal to mean Netflix was worried it couldn’t expand its current business, a theory co-CEO Greg Peters dismissed.
Warner Bros. stock was up about 5.2% midday in New York. It has almost doubled since reports of deal talks with Paramount emerged in September. Play Video
The news concludes a flurry of dealmaking over the past few months that began with a series of bids by Paramount. That prompted interest from Comcast Corp. and Netflix, who were both chasing just the studios and streaming business. All three submitted sweetened bids earlier this week, with Paramount ultimately offering $30 a share for all of Warner Bros. Discovery, arguing that its proposal offered a smoother path to regulatory approval. Netflix won out in the end although significant hurdles remain before the deal can close, which the company expects it can do in the next 18 months.
Paramount could still try to raise its bid, take its offer directly to shareholders or sue to try and block the Netflix deal. The company had no comment.
California Republican Darrell Issa wrote a note to US regulators objecting to any potential Netflix deal, saying it could result in harm to consumers. Netflix has argued that one of its biggest competitors, however, is Alphabet Inc.’s YouTube, and that bundling offerings could lower prices for subscribers. Netflix accounts for between 8% and 9% of TV viewing in the US each month, according to Nielsen. It accounts for closer to 20% or 25% of streaming consumption.
Analysts at Oppenheimer said platforms such as Reels, TikTok and YouTube competing for viewers’ time should help the deal pass antitrust review.
It was 15 years ago that Time Warner CEO Jeff Bewkes, who oversaw Warner Bros. and HBO, shrugged off the threat posed by Netflix, comparing the then fledgling company to the Albanian Army. As Netflix began to invest in original programming, Sarandos declared that Netflix wanted to become HBO before HBO figured out streaming.
Sarandos succeeded and Netflix led the streaming takeover of Hollywood while HBO struggled to respond to the rise of on demand viewing and the decline of cable. Bewkes agreed to sell Time Warner to AT&T in 2016, the beginning of a decade of turmoil for HBO and Warner Bros., storied brands that are about to have their fourth owner in a decade.
Warner Bros. put itself up for sale in October after receiving three acquisition offers from Paramount, which were rejected, opening the door for Netflix and Comcast. Peters said he didn’t see the logic of these big transactions at Bloomberg’s Screentime conference in October, but Sarandos privately pushed for the deal.
The bidding got contentious, with Paramount accusing Warner Bros. of operating an unfair process that favored Netflix. The Netflix offer topped Paramount’s when combining the money for the studio and streaming business with the estimated value of the networks. The two sides agreed to the deal Thursday night.
Under terms of the agreement, Warner Bros. shareholders will receive $23.25 in cash and $4.50 in Netflix common stock. Moelis & Co. is Netflix’s financial adviser. Wells Fargo is acting as an additional financial advisor and, along with BNP Paribas and HSBC Holdings, is providing $59 billion in debt financing, according to a regulatory filing, one of the largest ever loans of its kind. Allen & Co., JPMorgan Chase & Co. and Evercore are serving as financial advisers to Warner Bros. Discovery.
Netflix agreed to pay Warner Bros. a termination fee of $5.8 billion if the deal falls apart or fails to get regulatory approval. “We’re highly confident in the regulatory process,” Sarandos said Friday.
In addition to streaming overlap, regulators will also likely look at the impact on theatrical releases, which Netflix has traditionally eschewed in favor of prioritizing content on its platform.
Netflix said it will continue to release Warner Bros. movies in theaters and produce the studio’s TV shows for third parties — two major changes in how it does business. The company was a little short on details of exactly how it will integrate the different businesses, but Netflix said it expects to maintain Warner Bros.’ current operations and build on its strengths.
The deal will allow Netflix to “significantly expand” US production capacity and invest in original content, which will create jobs and strengthen the entertainment industry, the company said. The combination is also expected to create “at least $2 billion to $3 billion” in cost savings per year by the third year.
Warner Bros. Discovery CEO David Zaslav was the architect of combining Warner Bros. and Discovery in 2022, a deal he hoped would create a viable competitor to Netflix. But the company’s share price tanked in response to a series of public miscues and the continued decline of the cable network business.
While performance rebounded a bit over the last year, the company never quite became the streaming dynamo Zaslav envisioned. He’ll continue to run the company through its spinoff and sale. The two companies haven’t yet agreed on him having any role at Netflix.
The traditional TV business is in the midst of a major contraction as viewers shift to streaming, the world that Netflix dominates. In the most recent quarter, Warner Bros. cable TV networks division reported a 23% decline in revenue, as customers canceled their subscriptions and advertisers moved elsewhere.
In 2021, Mark Zuckerberg recast Facebook as Meta and declared the metaverse — a digital realm where people would work, socialize, and spend much of their lives — the company’s next great frontier. He framed it as the “successor to the mobile internet” and said Meta would be “metaverse-first.”
The hype wasn’t all him. Grayscale, the investment firm specializing in crypto, called the Metaverse a “trillion-dollar revenue opportunity.” Barbados even opened up an embassy in Decentraland, one of the worlds in the metaverse.
Five years later, that bet has become one of the most expensive misadventures in tech. Meta’s Reality Labs division has racked up more than $70 billion in losses since 2021, according to Bloomberg, burning through cash on blocky virtual environments, glitchy avatars, expensive headsets, and a user base of approximately 38 people as of 2022.
For many people, the problem is that the value proposition is unclear; the metaverse simply doesn’t yet deliver a must-have reason to ditch their phone or laptop. Despite years of investment, VR remains burdened by serious structural limitations, and for most users there’s simply not enough compelling content beyond niche gaming.
A 30% budget cut
Zuckerberg is now preparing to slash Reality Labs’ budget by as much as 30%, Bloomberg said. The cuts—which could translate to $4 billion to $6 billion in reduced spend—would hit everything from the Horizon Worlds virtual platform to the Quest hardware unit. Layoffs could come as early as January, though final decisions haven’t been made, according to Bloomberg.
The move follows a strategy meeting last month at Zuckerberg’s Hawaii compound, where he reviewed Meta’s 2026 budget and asked executives to find 10% cuts across the board, the report said. Reality Labs was told to go deeper. Competition in the broader VR market simply never took off the way Meta expected, one person said. The result: a division long viewed as a money sink is finally being reined in.
Wall Street cheered. Meta’s stock jumped more than 4% Thursday on the news, adding roughly $69 billion in market value.
“Smart move, just late,” Craig Huber of Huber Research told Reuters. Investors have been complaining for years that the metaverse effort was an expensive distraction, one that drained resources without producing meaningful revenue.
Metaverse out, AI in
Meta didn’t immediately respond to Fortune’s request for comment, but it insists it isn’t killing the metaverse outright. A spokesperson told the South China Morning Post that the company is “shifting some investment from Metaverse toward AI glasses and wearables,” pointing to momentum behind its Ray-Ban smart glasses, which Zuckerberg says have tripled in sales over the past year.
But there’s no avoiding the reality: AI is the new obsession, and the new money pit.
Meta expects to spend around $72 billion on AI this year, nearly matching everything it has lost on the metaverse since 2021. That includes massive outlays for data centers, model development, and new hardware. Investors are much more excited about AI burn than metaverse burn, but even they want clarity on how much Meta will ultimately be spending — and for how long.
Across tech, companies are evaluating anything that isn’t directly tied to AI. Apple is revamping its leadership structure, partially around AI concerns. Microsoft is rethinking the “economics of AI.” Amazon, Google, and Microsoft are pouring billions into cloud infrastructure to keep up with demand. Signs point to money-losing initiatives without a clear AI angle being on the chopping block, with Meta as a dramatic example.
On the company’s most recent earnings call, executives didn’t use the word “metaverse” once.
Politics8 years ago