Google told employees who want health benefits that they must allow a third-party AI healthcare tool to access their data, a move that has rankled some staff members.Google updated its policy after publication of this story.”Our intent was not reflected in the language on our HR site,” a Google spokesperson told Business Insider after publication. “We’ve clarified it to make clear that employees can choose to not share data, without any effect on their benefits enrollment.”The company announced this month that US-based employees who wish to sign up for health benefits through its parent company Alphabet in the coming enrollment period must grant access to AI-powered tools provided by Nayya, which offers personalized benefits recommendations, according to internal documents reviewed by Business Insider.A staffer will not be eligible for any health benefits if they decline to opt into Nayya’s tool, according to previous guidelines seen by Business Insider. Some staff members have asked leaders why they won’t have access to health benefits if they opt out of giving Nayya access to their data, internal communications show.Nayya’s tool allows employees to input information about their health and lifestyle, and provides recommendations on benefits to choose.”Nayya provides core health plan operating services to optimize your benefits usage, so Alphabet health plan participants can’t entirely opt out of third-party data sharing (as permitted under HIPAA),” read an internal resource page for employees.”To opt out of health supplier data sharing going forward, unenroll from the Alphabet-provided benefits during Open Enrollment or when you experience a qualified Family Status Change,” it continued.Google spokesperson Courtenay Mencini told Business Insider, prior to publication, that Nayya only has access to “standard” data about employees, such as demographic information, if they choose to opt in. From there, staffers can choose to make use of the tool and give it more data, or ignore it, she said.”This voluntary tool, which passed our internal security and privacy reviews, was added to help our employees better navigate our extensive healthcare benefit options,” Mencini told Business Insider. “Employees must opt-in to use the tool and share their own health information, as Google does not have access to it.”‘This is a very dark pattern’Employees posted messages on an internal Q&A site, asking why they must give potentially sensitive medical data to an outside tool if they want to receive medical coverage.

Related stories

Business Insider tells the innovative stories you want to know

Business Insider tells the innovative stories you want to know

“Why are we providing our medical claims to a third-party AI tool without a way to opt out?” read one of the questions submitted, which was viewed by Business Insider.”This is a very dark pattern,” read another post. “I cannot meaningfully consent to my data being shared with this company, and I do not want to consent in this manner.”Some employees also posted their concerns on Memegen, Google’s internal message board. One post read: “Consent for an optional feature like ‘benefits usage optimization’ is not meaningful if it’s coupled to a must-have feature like Google’s HEALTH PLANS! The word you’re thinking of is ‘coercive.'”A Nayya spokesperson said its tool allows employees who choose to participate to see how much of their deductible has been met and receive personalized plan recommendations. They also said that Google conducted a standard security and privacy review of Nayya’s product.”Nayya is required to protect health data maintained in accordance with HIPAA,” read an internal Google FAQ page about Nayya, which added that Nayya “will not share, rent, sell, or otherwise disclose” personally identifiable information it collects.Companies from Meta to Microsoft are increasingly incorporating AI tools into their workplaces. Google is pushing AI usage to increase employee productivity as well. Like Google, some companies, such as Salesforce and Walmart, have also rolled out AI-powered health benefits tools, like Included Health, to their workforces.Have something to share? Contact this reporter via email at hlangley@businessinsider.com or Signal at 628-228-1836. Use a personal email address and a non-work device; here’s our guide to sharing information securely.Update: October 8, 2025 — This story has been updated with an additional statement from Google. A spokesperson said the company changed the language on its internal site to clarify that employees can opt out of sharing data.

What used to take a team of designers and hours of effort can now be done with a prompt and a click. AI image generation tools are quietly becoming essential for professionals by streamlining workflows, unlocking new possibilities, and changing how we think about visual creation from the ground up. 

Adobe Firefly is one such AI image generation tool, integrated with the Creative Cloud system. According to the company, Adobe enables you to create images using tools like Generative Fill and also promotes ethical content creation through its Content Authenticity Initiative. 

Adobe’s generative tool integrates with Photoshop and Illustrator, enabling the creation of high-quality images. Apart from Adobe’s AI tools, which are made available, it also integrates with easy access to Adobe assets, such as music and sound libraries, used in various apps. Its ethical AI labelling makes Adobe a differentiator in the market, making it a commercially safe tool with no “copyright and IP violations,” said Prativa Mohapatra, vice president and managing director of Adobe India. 

Mohapatra added that while many AI models are as good as the prompts used for image generation, “Firefly has been trained on the images and content which belongs to Adobe, either [through] direct copyrights on it, or which is free” for Adobe. 

Yesterday, Adobe launched two versions of its text-to-image generative AI model, along with a range of new Firefly features and Creative Cloud app updates. The fourth-generation Firefly model also offers its customers two choices, one optimised for speed and efficiency, and the other for those requiring more tasks. 

To build a great model, you also need to consider its limitations. According to reports, Firefly lacks native vector generation, which is primarily focused on still images. Although it processes high-quality photos, the tool is also more expensive compared to other open-source or budget-friendly alternatives that professionals use. 

While acknowledging the “ease of use” tools, Mohapatra argues that Adobe does consider its limitations, specifically the lack of templates. However, she said that “it’s about timing,” and added that Adobe Express products are getting enhancements every day. 

Adobe also argues that the Firefly tool provides non-creatives a starting point, as “you can take an image from your photo library [to] start with something,” Mohapatra highlighted. 

Similarly, Canva’s AI tool generates images based on how well thought-out your prompts are, which could result in the creation of generic images. It also does not support image-to-video prompts, but Adobe’s AI tools are not just restricted to Firefly. “We started with images and augmented vectors. Video has already been launched. So I think it is a natural progression,” Mohapatra said.

However, it is well known that Adobe is used by professionals who already have an extensive understanding of how the tool works and are clear about their results. As for Canva, it is very beginner-friendly and provides a more versatile and budget-friendly tool. Some of Canva’s AI features are also open source, but their usage comes with limitations. 

Should You Choose Canva? 

Canva has been eyeing investments in AI for over seven years, much before the recent surge in AI design startups and tools. 

“AI is now integrated at every touch point that you have within Canva, from the home page where you might be suggested different designs…through to every step in designing, whether that’s removing a background or getting suggestions on copy,” Cameron Adams, co-founder of Canva, told AIM in an interview previously. 

While Canva is known for its templates and drag-and-drop design, the platform now offers the use of prompts for image generation, powered by Stable Diffusion and DALL-E models. Canva’s Magic Media tool uses third-party AI models to generate images. Canva is also known for its user-friendly interface, simple video editing and a marketplace for plugins. 

Canva features numerous AI tools powered by OpenAI’s algorithms, including DALL-E for image generation and GPT-like models for text generation. Canva has used these algorithms to integrate text-to-image capabilities, a text copy generator and a GenAI-powered design tool called Magic Design. 

Market in India

Canva’s widespread use across India as a people’s favourite image generation and editing tool has brought considerable success for the company. AIM reported that India has become Canva’s fourth-largest market, and Adams sees India as a powerhouse that has redefined digital design at a larger scale. 

Canva has also launched a Hindi website, which enables millions of users to communicate and engage with the tool in a language they are comfortable with. “It’s really important for us to bring a truly local product here. The team does an amazing job making sure we have the right content, the right product features, and the right languages that we’re speaking to people,” Adams said. 

Adams sees a huge potential for Canva in India, suggesting that it could become the number one market. “I think, given the current growth trajectory, being number one is just a question of time. India is a huge market,” said the co-founder. 

Interestingly, even Adobe views India as a significant market that regularly utilises AI generative tools. “In India, we are seeing a good optic for Firefly. So, more and more creative business houses are starting to use this,” Mohapatra added. 

However, she added that AI solutions “inherently build on a large hyperscaler environment. Scale at that base, it’s not an India-specific thing, and most of them are in the cloud, which is not constrained by the country.”

At its Adobe Max 2025 conference, the company launched two new text-to-image models, Image Model 4 and Image Model 4 Ultra, promising sharper visuals, better prompt accuracy, and increased realism. While Image Model 4 focuses on faster generation for basic illustrations and objects, Image Model 4 Ultra is designed for producing photorealistic portraits and complex scenes.

“These models bring higher fidelity and creative flexibility to our Firefly users,” Adobe said in a blog post.

The new models are now live for subscribers through the Firefly app, alongside powerful filters, style matching, and composition tools.

Firefly Goes Mobile and Welcomes Third-Party AI

The launch also included the first-ever Firefly mobile app, giving users on-the-go access to Adobe’s creative AI suite. Adobe is further opening Firefly to third-party models, with integrations for OpenAI, Google’s Veo 2, and Flux 1.1 Pro now live, and future partnerships in the works with fal.ai, Ideogram, Luma, Pika, and Runway.

A new Vector Model was also unveiled, allowing designers to generate editable vector artwork, such as logos, packaging, and scenes, using natural language prompts.

Another notable debut is Firefly Boards—a public beta tool for storyboard and concept creation. Previously codenamed Project Concept, it enables users to explore visual ideas with text-to-image generation, AI-powered refinements, and mood board editing, all within one interface.

Firefly Video Model Goes Public

After being previewed last year, the Firefly Video model is now generally available. It supports text and image-based video generation, useful for creating custom b-roll, background visuals, and stylised edits—all driven by AI.

Adobe Express is also seeing a wave of AI-powered upgrades for video editors. A new Clip Maker tool transforms long videos into shareable clips, using AI to detect key moments and add captions or reframing.

Other additions include:

    •    Enhance Speech: Removes background noise for better audio clarity    •    Video Self-Record: Enables self-filming inside Adobe Express    •    Drop Zone and Scene View: For batch clip editing and timeline rearrangement    •    Dynamic Animation: Turns still images into motion-rich content

Users can now export directly to Vimeo, and enterprise customers benefit from a Generate Similar feature that builds on-brand variations from a reference image. Over 30 new visual filters are also rolling out across the platform.

Firefly’s entry-level subscription starts at $9.99/month (approx. ₹852), with plans tailored for individuals, teams, and students.

As Adobe steps deeper into the AI-driven creative ecosystem, the suite of updates across Firefly and Express marks a significant push to make generative tools more accessible, intuitive, and professional-grade, whether on desktop or mobile.

There was, however, one thing that Netflix did that raised some eyebrows, using AI to recreate Gabby Petito’s voice for text messages and letters she wrote, trying to have it be in her own words rather than having another actress read them, as we usually see in programs like this.

There is a significant anti-AI movement in many circles, and on social media, there were a lot of negative reactions to this decision, among other things, calling it “disrespectful” to Petito and her memory. But there’s a group who did not agree with this, Petito’s family.

In talking about the decision to use the tech to US Weekly, a Netflix producer had this to say:

“We reached out to the family to get their blessing and then we worked diligently to represent it in exactly how it was written,” executive producer Michael Gasparro. “That allowed you to hear it through her own words…We thought it was really important to bring that to life. At the end of the day, we wanted to tell the story as much through Gabby as possible. It’s her story.”

It’s a unique case as this tech has not been this advanced in the past, in this instance, since the family gave permission for this to be done, that complicates perceptions of whether this may be “okay” or not.

Hollywood has dealt with both of these issues before, albeit usually separately. This year’s Best Picture nominee, The Brutalist, got flak for using AI to make accents in the film more accurate. Elsewhere, we see Hollywood recreate deceased actors for films using old footage or pure VFX (this has happened in Star Wars, Alien Romulus, and others as of late).

This happening in this way is sort of a first-of-its kind situation in terms of how advanced the tech has gotten, and again, the amount of footage from Petito herself no doubt contributed to the ability to do this. Watching the show, you can definitely hear how it sounds somewhat “unnatural,” but you sort of get used to it as time goes on, and it just becomes another part of the show eventually. Though no doubt opinions on its use here will continue to differ.
Follow me on Twitter, YouTube, Bluesky and Instagram.
Pick up my sci-fi novels the Herokiller series and The Earthborn Trilogy.

Instead, Humane is winding down support for the Pin, pausing orders and encouraging users to download their media, according to a note shared by the company.

The Pin will function as normal until February 28, at which point it’ll lose most of its features, including calling, messaging, connecting to the cloud and its AI-powered shenanigans.

Some Pin users will be eligible for refunds, but anybody who copped the wearable device prior to November 15, 2024 is out in the cold. Refund requests must be submitted by February 27.

So while Humane’s AI division, including its founders and employees, will get rolled into HP as part of the $116 million acquisition deal, a chunk of consumers will be pretty much left with the husk of a device once touted as the next big thing in tech.
Founders Imran Chaudhri and Bethany Bongiorno are expected to form a new division at HP, where they’ll oversee the integration of AI within the company’s computers, printers and conferencing tools, Bloomberg reports.

Humane, which received $230 million in funding from big shots like OpenAI’s Sam Altman, Salesforce’s Marc Benioff, SoftBank, Qualcomm, Microsoft, LG and Volvo, has reportedly been seeking a buyer since May 2024, initially hoping to offload its business for a sum in the range between $750 million to $1 billion.

Well, the final number is $116 million — the price of disappointed early adopters not included.
Death By YouTube?
Humane’s flagship AI Pin, which began shipping less than a year ago in April, received unfavorable reviews across the board.

Brownlee’s reflections on the wearable, summarized in a now-infamous video titled “The Worst Product I’ve Ever Reviewed… For Now” published in April 2024, drew significant controversy at the time.
Although the YouTuber was praised for sharing his unfettered thoughts on the device, naysayers expressed concern that his negative feedback could hurt a company trying to do something new in an otherwise increasingly stale tech ecosystem.
In hindsight, Brownlee had every right to be skeptical — especially now that we know the ones most affected will be the people that believed in Humane’s vision the most and backed it with their hard-earned cash.
Humane’s AI Pin was always a long shot. The proliferation of AI and the acceleration of its adoption into phones and other consumer devices made an already cramped space even smaller for the startup. Who could’ve seen this coming? Brownlee did.

Today, former OpenAI CTO Mira Murati announced her new venture: Thinking Machine Labs, a public benefit corporation that aims to build accessible and broadly capable artificial intelligence systems. After leaving AI juggernaut OpenAI last September, Murati has brought together a team of engineers and researchers who have worked at buzzy startups like Character AI, Mistral and unsurprisingly, OpenAI. Additionally, Thinking Machine Labs said it will publish its technical blog posts, code and papers and collaborate with the broader community, indicating it plans to open source its work.

Now let’s get into the headlines.

BIG PLAYS
The cluttered world of AI reasoning models just got its newest addition. On Monday Elon Musk’s AI company, xAI launched a new AI model called Grok-3 that can process and answer complex questions across domains like science and math. In a livestream on X, Musk said the company’s mission is to “understand the universe…to figure out what’s going on, where are the aliens and what’s the meaning of life.”

The billionaire claims the model is built using 10 times more compute than Grok 2, likely from its “gigafactory of compute” in Memphis, and that it has been trained on public data from sources including social media platform X and legal documents. xAI also rolled out an AI-powered search engine called DeepSearch. OpenAI cofounder and former Tesla executive Andrej Karpathy, who tested the model, says Grok-3’s capabilities are largely on par with OpenAI’s best models but gets some questions wrong. Others have noted the model is lacking in its coding abilities compared to others. Grok-3 has yet not been independently evaluated and is only available to paying users.

ETHICS + LAW
Condé Nast, Vox, The Atlantic and a group of publishers have sued $5.5 billion-valued AI company Cohere for copyright and trademark violations. (Forbes is part of the group suing Cohere.) The lawsuit alleges that the Canadian AI startup scraped 4,000 copyrighted articles from the internet and used them to train its family of large language models called Command, which reproduced sections or entire works (at times word for word), allowing users to get information without visiting the publishers’ websites. It’s not the first time an AI company has faced publishers’ scrutiny. Last year, AI search engine Perplexity came under fire for republishing copyright works from multiple publications including Forbes. (In response, Forbes sent a cease and desist letter to Perplexity, accusing it of copyright infringement.)
AI DEALS OF THE WEEK
Humanoid robotics company Figure AI is in talks to raise $1.5 billion in venture capital at an eye-popping $39.5 billion valuation, Bloomberg reported. The news comes as the company is reportedly in talks with Meta to make robots for household chores.
AI legal company Luminance, which helps customers like AMD and National Grid generate, negotiate and analyze contracts, has raised $75 million in series C funding.
Chip startup Encharge AI has raised $100 million in a series B funding led by Tiger Global. CEO Naveen Verma started the company out of a lab in Princeton, where he worked on designing architecture for hardware that would help run large language models more compute and energy efficiently. Verma says the chips allow AI models to run locally on devices such as personal computers.
DEEP DIVE
Elon Musk’s surprise bid for the nonprofit controlling artificial intelligence behemoth OpenAI did exactly what he wanted it to. Announced as OpenAI CEO Sam Altman and other business and world leaders convened in Paris for a global AI summit, the unsolicited $97.4 billion offer for the nonprofit refocused the world’s attention on Musk and his efforts to block OpenAI’s transition to a for-profit company.
An irked Altman quickly dismissed Musk’s offer and sources close to OpenAI say it’s hard to imagine it going anywhere. But even if that’s the case, Musk has likely caused a headache for Altman, who is orchestrating the company’s transition to a for-profit venture. He’s attempted to forcefully raise the nonprofit price – which would make it harder for OpenAI to justify paying anything less.
Musk’s bid is the first hard number that values the nonprofit that controls OpenAI; that entity has to be bought out and become a minority shareholder for OpenAI to successfully transition to a for-profit company. Previously, The Information had reported the nonprofit was worth around $40 billion, citing a 25% stake and the company’s valuation at time. But with his $97.4 billion bid, Musk has backed Altman into a corner; now, as a board member, Altman faces pressure to sell the nonprofit for at least what Musk is asking. If he were to sell for anything less, it’d be a bad look, making it seem like he’s lowballing his own company to reduce share dilution.
“The important part here is that if [the board] doesn’t take it, which they almost certainly won’t, then they’ve made clear that they think the assets Musk is trying to buy are worth more than $97 billion,” a person familiar with the company told Forbes. “So if the for-profit tries to buy them later, the nonprofit will have to get more than that — otherwise the board is likely in breach of their fiduciary duties.”
Read the full story on Forbes.
MODEL BEHAVIOR
Generative AI is making it easier for fraudsters to carry out romance scams at scale, Wired reported. AI chatbots are being used to generate hundreds of deceptive scripts and generate fully fake profiles on dating apps. AI has already made a foray into the dating world. Last year, we wrote about a man who programmed ChatGPT to reply to his matches on Tinder and set up dates for him.