This story is over 5 years old.


Sesame Messaging App Uses Digital Assistants to Protect Privacy

Sesame boasts about end-to-end encryption, data ownership, and digital helpers that make sure your private data is always under your control. Now it just has to prove that it can be trusted with all of that information.
Image: Feed, Inc.

It's hard to know what companies are doing with our data. Countless tools are used to monitor web traffic, activate a smartphone's microphone to listen for television ads, and gather other personal information that marketing companies can use to inform the advertisements they show. Now a Silicon Valley startup called Feed wants its "personal cognitive assistants," which don't currently have a user interface like Siri or Cortana but instead operate behind the scenes, to help give people some control over what data they share, and with whom. And it's released an iPhone messaging app called Sesame to help people understand what these not-quite-artificially-intelligent helpers can do.


Sesame works like many other messaging apps: People sign up for the service, communicate with their friends, and then put the app in a folder with all the other messaging tools they have installed. But underneath that veneer of sameness is a tool that is supposed to give its users the ability to decide how their messages, images, and other information is handled, even after that data has made it to someone else's phone. Right now this tool is simple—it stops messages from being forwarded or photos from being saved, for example—but it will continue to evolve.

All of this relies on the cognitive assistants Feed has created. These vigilant watchers keep track of their assigned human's data, find out how it's supposed to be handled, and do their best to make sure everything is as it should be. So if someone tells Sesame they don't want their messages to be forwarded, those messages won't be forwarded. And if they say a photo isn't supposed to be saved, eventually these assistants will be able to detect if it's screenshotted, pass that information along, and nag the screenshotter to delete that photo from their phone.

The idea is that getting people to trust Sesame with their texts, sexts, and other personal communications will eventually allow Feed to access even more information. "Sesame is basically our proof of concept," says Feed co-founder and chief executive Mitch Ahlenius. If consumers trust this app with their information, he says, they'll move forward with their plans to use these cognitive assistants in ways that don't involve messaging. To belabor the obvious pun: Feed is hoping that Sesame will open the door for bigger and more lucrative things.


Ahlenius describes a scenario in which someone has dropped their phone. A cognitive assistant could notice this, send the information to case manufacturers, and haggle on a discount for a phone case bearing the logo of its owner's favorite sports team. Or it could do none of these things: Feed promises that people will be able to choose what information is collected, with whom it's shared, and how it's handled. These decisions won't be permanent, either—Ahlenius says that someone could choose to stop sharing their data with the outside world at any time.

But right now Feed just has Sesame, which will focus on giving people control over their texts. And some aren't convinced the app is as secure as it claims. "You still have no idea how much data you're giving, where that data's being stored, how it's being used," says Privacy International technology officer Christopher Weatherhead. "This isn't just true of this app. This is true of all telecommunications or internet-accessible devices. You just don't know where stuff is being sent and who's got access to it." He adds that unless an app is open source and has been subjected to independent audits, people can't know if Sesame makes good on all its promises.

Ahlenius says that Feed plans to have Sesame audited "shortly" and that the company is "the biggest [believer] in independent audits" even though it's "focusing [its] budget on development." This means that many of the company's claims—from its promise of end-to-end encryption to users maintaining agency over their data—must currently be taken on faith. That could be a hard sell given the renewed interest in secure tools due to the Apple-FBI fight, and considering Feed's hopes to eventually handle far more personal data than Sesame currently manages.

In theory, Sesame hits all the right notes. Its app is relatively easy to use, the company's promises about data are preferable to the collect-and-share-it-all attitude taken by many tech companies, and every messaging service should promise end-to-end encryption of its users' communications. "A lot of this is marketing," Weatherhead says. "It's hyperbole rather than really provable security. We have no idea what our devices send and how they communicate and what data they store. It's just impossible to know."

Welcome to the new age of privacy theater.