2018: A year of creepy AF smart home headlines

2018 is just about over, and it's common for tech reporters to dig back into their beats to try and sum up the year's news. And, as the CNET Smart Home team took a look back for our own year in review, there was one takeaway we just couldn't escape:

The smart home was... kinda creepy this year.

To wit, there were evil cackles from Alexa, fake phone calls from Google Assistant, and concerns a-plenty about the connected cameras and microphones filling our homes. That's not to mention the well-founded fears about the mass amount of data that these devices are asking to be trusted with (looking right at you, Facebook Portal. No thanks.)

So, yeah, if the smart home had you a bit creeped out in 2018, I can't say I blame you. Here's a look back at the stories that probably played a role.

Microphone mishaps (no laughing matter)

Things typically get a bit quiet in tech during the months following January's big CES expo in Las Vegas. This year, a mysterious and disconcerting bout of laughter echoed out of the silence.

The cackle in question came from Alexa, when users of Amazon's popular voice assistant began sharing clips of eerie, unprompted laughter emitting from their Echo devices. Once the phenomenon began to trend on Twitter, Amazon confirmed the issue and told us it was looking into a fix.

"In rare circumstances, Alexa can mistakenly hear the phrase 'Alexa, laugh,'" the company ultimately explained. "We are changing that phrase to be 'Alexa, can you laugh?' which is less likely to have false positives, and we are disabling the short utterance 'Alexa, laugh.' We are also changing Alexa's response from simply laughter to 'Sure, I can laugh' followed by laughter."

In another Alexa headache this May, a family in Oregon claimed that their Echo device recorded audio of a private conversation and sent it out to a random contact without warning. Amazon's explanation? It was the Alexa equivalent of a butt-dial.

"[The] Echo woke up due to a word in background conversation sounding like 'Alexa,'" an Amazon spokesperson told CNET. "Then, the subsequent conversation was heard as a 'send message' request. At which point, Alexa said out loud, 'To whom?' At which point, the background conversation was interpreted as a name in the customer's contact list. Alexa then asked out loud, '[contact name], right?' Alexa then interpreted background conversation as 'right.'"

"As unlikely as this string of events is, we are evaluating options to make this case even less likely," Amazon added.

Still, accidents happen. Just today, Reuters reports that Amazon mistakenly sent the wrong files to an Alexa user in Germany who had requested his Echo recordings. As a result, he was able to download and access 1,700 recordings from another user's household. Per Reuters, Amazon called this an isolated incident, and the result of human error.

Assistant anxiety

Amazon also found itself navigating potential legal battles over user privacy when Alexa became an apparent witness to not one, but two separate cases of homicide. The first came to a conclusion in March when Amazon agreed to hand over the Alexa audio recordings of a user in Bentonville, Arkansas, who stood accused of of first-degree murder, but only after that user consented to the release.

More recently, a judge in New Hampshire ordered Amazon to hand over the Alexa audio of a user accused of two counts of first degree murder. For now, Amazon is fighting the ruling.

"Amazon will not release customer information without a valid and binding legal demand properly served on us," an Amazon spokesperson told CNET. "Amazon objects to overbroad or otherwise inappropriate demands as a matter of course."   

Concerns about silent, "ultrasonic" audio attacks also raised our collective assistant anxiety this year, with researchers claiming that the most popular voice platforms were vulnerable to audio cues at frequencies too high for humans to hear. Researchers at UC Berkeley claim they were even able to fool Mozilla's open-source DeepSpeech voice-to-text engine by hiding ultrasonic audio cues within brief snippets of music.

Maybe most disconcerting -- to date, none of the major tech companies responsible for these voice platforms have denied that attacks like these are possible.

Enlarge Image

Big moonshots, big questions

2018 saw some notable new ideas about where smart home tech may be headed -- and some of those ideas raised a lot of questions.

Let's start with Google Duplex, the search giant's effort to let the artificially intelligent Google Assistant make phone calls on your behalf. Google touted the feature as a way for the Assistant to book things like haircuts and dinner reservations -- and the demo was pretty mind-blowing. Watch for yourself below:

Watch this: Robot or human? Google Assistant will leave you guessing

04:25

Google's efforts were impressive, but Duplex didn't sit right with all of us. "In this age of disinformation, where fake news thrives and the public has trust issues with technology, Google designed a machine that can deceive humans," wrote CNET's Bridget Carey, who also voiced concerns over the fact that Google's Assistant never identified itself as a robot. "Gosh, what could go wrong?"

Two days after that editorial, Google told CNET explicitly that it will launch Duplex with "disclosure built-in" for better transparency.

Then there's Amazon Key, which looks to leverage smart lock tech at your front door to let delivery people drop Amazon packages off inside your home. The Amazon Key Home Kit even includes an Amazon Cloud Cam to let you watch the delivery in real time right from your phone (more on cameras in just a bit, by the way).

The idea of granting a stranger authorized access to your home was unnerving for many of us -- but in the end, Amazon Key wasn't nearly as creepy as we expected. In fact, the most we ever saw of the delivery people was an arm reaching inside the cracked door to drop a package off just inside. From Megan Wollerton's recap:

"While I started out hesitant about in-home delivery, I'd be comfortable enough to use it going forward (and to recommend it to frequent Amazon customers) -- as long as I can keep an eye on whoever's delivering the packages from my phone."

Camera creep

2018 might be the year when the mainstream smart home opened its eyes, with a flood of new camera-equipped products and services designed to help us keep watch over our homes or stay in touch with loved ones. Early in the year, for instance, Amazon bought the home security startup Ring and its popular video doorbell. Now, a new app for Ring users called Neighbors lets you share, view and comment on local crime activity, complete with video clips from Ring cameras and doorbells in your community.

That's an appealing pitch to some, but consider the company's recent patent applications focused on facial recognition. Each considers ways to use Ring cameras to identify "suspicious" people (convicted felons, sex offenders, etc.), then automatically alert law enforcement. That raised fresh surveillance state concerns from the American Civil Liberties Union.

"Amazon is dreaming of a dangerous future," the ACLU's Jacob Snow said in a statement, "with its technology at the center of a massive decentralized surveillance network, running real-time facial recognition on members of the public using cameras installed in people's doorbells."

We expect to see more uses of facial recognition technology in the coming year, including from cameras that are already equipped with the feature, like Google's Nest Hello video doorbell. In other words, don't expect this issue to fade from sight.

And it's not just the front door -- from night-vision security cameras to connected baby monitors, people are parking cameras inside their homes, too. Doing so might require you to wrap your head around some potential privacy vulnerabilities. For instance, earlier this year, researchers from Kaspersky Lab warned that hackers could turn your own cameras against you by spying on you, or by fooling you with a duplicate, "cloned" feed. Your next vacation rental might have security cameras inside, too, which raises concerns about whether or not you could be filmed without realizing it during a weekend getaway.  

That brings us to smart displays, a new smart home category that promises to bring even more cameras into people's kitchens and living rooms. Amazon was first into the space with the Echo Show and Echo Spot, and Google soon followed suit with a suite of its own Google Assistant-powered touchscreens, including the Lenovo Smart Display and the JBL Link View. Tellingly, Google opted not to include a camera in its flagship, first-party smart display, the Google Home Hub.

Of course, there's another smart display worth mentioning. In fact, it gets the next section of this post all to itself.

Smart home hubris from Facebook

There was a collective "you've gotta be f***ing kidding me" from many of us who write about tech when Facebook, in the midst of scandal after scandal after scandal over the misuse of user data, unveiled the new Facebook Portal in-home video chatting devices. Because sure, why not let a transparency-challenged company that's been -- at best -- wildly irresponsible with user data bring person-tracking cameras and always-listening microphones into your living room? They promise they'll behave!

And hey, amid escalating scrutiny, perhaps they will. "We have a responsibility to protect your data," Facebook CEO Mark Zuckerberg wrote in a March 21 statement following the company's Cambridge Analytica data-mining scandal. "And if we can't then we don't deserve to serve you."

Just keep in mind that since then, Zuckerberg's company has come under continued fire for not doing enough to deal with a flood of fake news and Russian trolls on its platform during the 2016 US election. In July of this year, the company admitted to sharing user information with hardware and software partners even after claiming to discontinue the practice in 2015. Just this week, the New York Times reported that Facebook used "special arrangements" like those even more than the company initially disclosed -- namely, to give Microsoft's Bing search engine access to the names of all Facebook users' friends without consent, and to allow Netflix and Spotify to read Facebook users' private messages.

Netflix denies that it ever accessed people's private messages on Facebook, or asked for the ability to do so. Spotify didn't immediately return a request for comment.

At any rate, we are well past the point of taking Facebook's word for anything when it comes to protecting your private data. But, if early market indications showing lots of interest in smart displays this holiday buying season are correct, that's exactly what some people might be doing. That, or they've been drawn in by Facebook's admittedly slick user interface -- enough so to shrug off some extremely valid privacy concerns.

After the year we've had, maybe that's what creeps me out the most.