Meta will charge for access to advanced AI features

Listen to the article 6 min

This audio is auto-generated. Please let us know if you have feedback.

The signs are not looking great for Meta’s expanded artificial intelligence ambitions. The company is now looking to add more charges to access its advanced AI tools as it works to offset its massive investment into infrastructure and development.

Last week, Meta employee David Woodland posted an overview of Meta’s latest AI advances in an update on X, which included this note at the bottom of the list:

“We’re gradually rolling out Meta One to unlock premium features across Instagram, Facebook, and WhatsApp, as well as more access to AI capabilities across our apps and AI glasses. Subscribing to Meta One Premium plan will give device owners expanded access to conversation focus and premium device support.”

Meta officially announced the update yesterday, which means that Meta AI glasses owners will now have to pay a monthly fee to get maximum use out of the device.

Of course, conversation focus, which amplifies the voices of people that you’re speaking to in noisy environments, is just one element of the device. But even so, paying a monthly subscription to unlock the functionality of a purchased device could feel like an overreach for consumers, while the update may also point to future paid features in the app. 

According to previous reporting from CNBC, the Meta One Premium package is priced at $19.99 per month. With this, Meta AI owners would get unlimited access to conversation focus, while those who don’t pay for the add-on subscription will only have the feature available for three hours each month. 

In addition, as noted by Woodland, subscribing to Meta One Premium will also give users access to improved customer support for the device. That also seems like it should be a standard part of the service, but customer service has clearly never been a part of Meta’s prime focus.

Though that approach could become a bigger problem in the hardware arena.

That said, the updated charges for AI elements aren’t a huge surprise, given that Meta has been warning for many months that it may eventually need to charge for some AI features, due to the rising cost of its AI development.

In May 2025, when previewing upcoming AI-powered tools for the Edits app, Instagram Chief Adam Mosseri said that Meta may eventually have to charge a fee for the tool, or add a specific charge to use its advanced AI features.

Meta also recently announced a new set of add-on subscription offerings across its apps. That upcharge also includes fees for coming AI elements.

The updated Meta AI subscription charges also come as Meta considers moving into cloud storage, and charging other AI projects for access to additional compute. 

In combination, the expansion of fees for its AI offerings is beginning to suggest that the company’s massive AI investment may not be coming together as it had hoped, and that its projections about the potential of the technology may have oversold its value.

Indeed, Meta has even had to walk back its forecasts regarding how AI will impact its own workforce. Business Insider reported that Meta CTO Andrew Bosworth recently issued an apology to staff for the “atrocious” way the company had handled recent layoffs and restructuring.

Meta has cut thousands of jobs in 2026, with a significant portion of those roles seemingly replaced by AI systems. Meta CEO Mark Zuckerberg has repeatedly touted the potential of AI on this front, and has claimed that, eventually, most engineering roles, within Meta and other organizations, will be able to be replaced by AI agents.  

Meta’s job cuts, then, seemed like both a means to offset its increasing infrastructure spending, and a proof-of-concept project. Though some teething problems suggest that this process may not be as straightforward as Zuckerberg had envisioned.

For example, more than 20,000 Instagram accounts were recently hijacked by hackers who used an exploit in Meta’s AI-powered account support system to trick the bot into granting them access. Meta has already replaced many moderation and support staff members with AI chatbots, and the company recently announced plans to replace almost all of its moderation workforce with AI by the end of this year.

Given the Instagram hack, and the expanded challenges in patching AI prompt manipulation, that could open the company up to more vulnerabilities. Meanwhile, as reported by Reuters, Meta’s AI chatbots have engaged in sexualized chats with kids, adding to the list of potential problems caused by empowering AI chatbots without human oversight.

Add to this the fact that consumer and market sentiment on AI is cooling. Social media users have criticized the latest AI tools being foisted upon them by enthusiastic platforms, and reports suggests that many businesses are not seeing the gains that were promised with AI after implementation.

A key point of contention in terms of AI-enabled efficiency is in actual implementation. AI vendors are going to companies and outlining idealistic depictions of an AI-assisted workflow, where creators and professionals can get more done, more efficiently, by utilizing AI to complement their processes.

And that might be true, but it also overlooks the fact that the vast majority of people are lazy, or at the least opportunistic, and as such, offering them a means to automate tasks instead of manually doing them means that many will take advantage of exactly that. As a result, AI tools aren’t always being used as prescribed, with the requisite cross-checking and quality checking to ensure their outputs are correct.

Compounding this element is the fact that with proper cross-checking and analysis, it often ends up taking just as long to produce work via AI as it would to just do the whole thing. And with workers under pressure to utilize the latest AI tools to save time, in order justify the company’s investment, that means organizations are increasingly seeing poorer quality work, with more unchecked, fully AI-generated content being submitted by staff.

The actual benefits of AI, then, may end up being more limited than what has initially been pitched. Meta could now be seeing that play out in application, leading it to seek out expanded monetization pathways.