As part of its annual Connect showcase this week, Meta unveiled a range of new products that will enable people to further integrate the company’s artificial intelligence agents into their everyday lives, by having them tune in to their interactions at all times, via an expanding range of wearable devices.
But can Meta be trusted to listen in on people’s daily conversations, and use those data inputs in a responsible way?
This is a key question in the company’s broader agentic AI push, that while Meta’s AI agents may be able to undertake a range of tasks on users’ behalf, and potentially deliver value and utility, that will require a level of trust in the company that many will likely be hesitant to allow.
Which is only going to expand as Meta rolls out more always-on AI assistant products.
Meta is already working to address these concerns. As part of its Connect announcements, it shared details of its new Private Processing approach for Meta AI, which will process user information gathered by its devices inside “confidential virtual machines,” which the company said will ensure that not even Meta itself can access that information.
It also announced that users will now be able to opt-out of having their visual data from AI glasses used for model training, while it also unveiled a camera-less version of its AI glasses, in order to address concerns about unauthorized recording, while still facilitating AI connection.
Other reveals included a new toy-like charm device for Meta AI access, called Muse Charm, which will provide another, non-camera-equipped way for people to maintain connection with their Muse AI agents.
The most interesting thing about the charm is the disarming nature of it, that Meta is using cute characters and fun devices as a means to drive adoption of an always-on AI recording device.
Because again, Meta needs users’ data, all of it, in order to make its agentic AI push work.
As explained by Meta: “The more you use an AI assistant, the more useful it gets as it learns your style, preferences and context. Wearable AI assistants will help you in similar ways, including with everyday life. To do that, they need to know you and your context. That can include connecting ideas across days or weeks, and working proactively in the background to get things done for you without requiring rework from you.”
Meta wants to be listening in at all times, but the level of information sharing that will be required is massive, and given the company’s past history on this front, it will likely be a bridge too far.
Unless Meta can demonstrate real utility, or facilitate a new trend, like through Meta Charms, that supersedes privacy concerns.
We’ve already seen this with social media, that people have willingly continued to share their sensitive personal data with social media platforms, despite various violations of that trust, and concerns around data misuse, because social media connection has become so ubiquitous, and arguably a vital element of modern interaction.
Meta is hoping that it can counter concerns with these new products, and reassure enough users, while also looking to win them over with utility and cuteness. But again, Meta has repeatedly shown a willingness to push forward with innovations first, then address concerns in retrospect, once the damage has been done.
On balance, that seems likely to be the case once again with AI agents.
