“Instead of waiting until someone breaks down your door and then calling the police, we want to detect danger before it happens,” said Christian Cerda, CEO of SimpliSafe.

SimpliSafe pioneered inexpensive security systems that do not require expensive wiring and can be installed by homeowners. The company now hopes to build a system intelligent enough to detect potential burglars. For now, the system still relies on human assistance to decide whether someone is friend or foe. Officers are warned not to use unfair criteria like race when making such decisions. But Cerda hopes to one day create a version capable of making such decisions on its own. And that possibility makes privacy advocates a little nervous.

“I don’t want a robot to decide that one person’s behavior is dangerous and another person’s is not,” said Adam Schwartz, director of privacy litigation at the Electronic Frontier Foundation, an organization defending digital civil liberties.

“Stupid” security cameras like Amazon’s Ring Doorbell cameras have previously been criticized for posing a privacy threat. Under pressure from civil liberties groups, Amazon said in January that it would no longer allow police departments to request video footage from Ring camera owners for use in criminal investigations.

The SimpliSafe system raises questions about whether it is permissible to perform facial recognition scans without the subject’s permission and whether an AI is intelligent enough to correctly identify possible criminal activity.

The new SimpliSafe system captures video of people approaching the home and alerts a human agent at the company’s alarm monitoring center. The camera’s AI system can learn to recognize the faces of family members and friends. If the visitor has a familiar face, the SimpliSafe agent is asked to leave.

But for anyone else, the AI ​​generates a textual description of the person. For example, it might be a report that “an unknown person wearing a green shirt and blue pants was detected by the front door camera at 10:15 p.m..”

The agent can watch live video from the camera and speak to the visitor through the camera’s built-in speaker and microphone. The agent would say something like, “This is SimpliSafe. Your visit is recorded on video. Can I help you?” If the visitor is delivering a package, no problem. If it’s a criminal, the officer can call the police.

This feature is built into the company’s new outdoor security camera, priced at $200. Users will also pay $80 per month to have humans monitor the camera 24 hours a day, or $50 for nighttime monitoring only.

“I would say what they’re doing is pretty advanced,” said Elizabeth Parks, president of market research firm Parks Associates. She said it was one of several efforts by home security companies to integrate AI technologies such as facial recognition into their products. “The major players and new entrants are all looking at the benefits of this machine learning,” she said.

SimpliSafe already makes an indoor camera that allows officers to warn thieves after detecting a break-in to the property. The new outdoor camera is supposed to deter them before they break a window or break down a door.

According to Hooman Shahidi, director of products at SimpliSafe, many thieves are not afraid of burglar alarms. “When people break into houses, they hear the siren, and they say, yeah, of course there’s a siren or whatever,” Shahidi said. “I have 20 minutes before someone arrives.

Indeed, in some American cities, the police can take an hour or more to respond to an emergency call. Shahidi said part of the reason is that about 90 percent of burglar alarm calls are false alarms. Knowing this, the police are often slow to react. This could discourage people from buying an alarm system – why bother if help comes too late? SimpliSafe therefore hopes to win over more customers with a system that tries to scare away thieves before calling for reinforcements.

By combining AI with a human security officer, the SimpliSafe system minimizes the risk of a computer calling the police in response to an innocent visit. But the system still raises some civil liberties concerns.

For example, the automatic face matching feature will not be available to SimpliSafe customers in Illinois, Texas, and Portland, Oregon. These jurisdictions have enacted laws that prohibit the collection of facial recognition data without the individual’s permission. Mapping the faces of anyone who shows up could violate these laws.

There are also concerns that SimpliSafe’s new system is only the beginning. How long before it uses AI instead of humans to decide who poses a threat? In an email, Cerda said that while the system doesn’t currently do this, he has made it clear that SimpliSafe is moving in that direction.

“After our launch period and as we gain scale, we will continue to refine this technology, further train and validate our models, and focus on specific cases where we are confident not to introduce bias or other errors,” Cerda wrote. “At this point, we will use the tool to filter or even take action on an event; this is essential in our roadmap to deliver this service at scale and at lower cost.

Cerda declined to provide further details on what type of action the AI ​​might take. But it’s possible that a future SimpliSafe system could respond to an unidentified visitor using AI alone. The computer can decide to issue a verbal warning, or even call the police, without human intervention as backup.

Companies are already using AI facial expression analysis to gauge people’s reactions to television commercials or political campaign ads. And self-driving car developers are working on AI systems that would assess the body language of nearby pedestrians, to predict whether they will enter the street.

This has raised concerns in some quarters that a company like SimpliSafe could eventually use these methods to automatically assess criminal intent. And that’s what worries Schwartz.

“All the biases and irrationality of human life are fed into AI and reproduced by AI,” Schwartz said. He fears that racial and class biases built into the software could harm innocent people.

Cerda admits that this is a legitimate concern, at least until SimpliSafe builds better AI models. This is why the system will maintain control over humans – for now.

Hiawatha Bray can be contacted at [email protected]. Follow him @GlobeTechLab.