The Dark Side of Voice Assistants
Mon, April 19, 2021

The Dark Side of Voice Assistants

It was estimated that by 2023, there will be more than eight billion digital voice assistants across the world / Photo by prykhodov via 123RF

 

UK-based analysts at Juniper Research reported that the use of voice assistants is set to triple over the next few years. They estimated that by 2023, there will be more than eight billion digital voice assistants across the world, which is a substantial increase from the 2.5 billion in use at the end of last year. This shows that the technology has become a routine part of everyday life for many people.

Ironically, these voice assistants like Siri, Google Assistant, and Cortana weren’t making much of a difference when they were first introduced. Users then thought that they weren’t delivering much value to them, primarily due to their error rate and the limitations of their artificial intelligence. Of course, as we all know by now, their inability to provide helpful, rich, and contextually influenced answers eventually improved tremendously in recent years. There are now more data and better tuned AI algorithms on voice assistants.

The rise of voice assistants, specifically Alexa, Google Assistant, and Siri, is also influencing industries. Canalys, an independent analyst company that strives to guide clients on the future of the technology industry and to think beyond the business models of the past, reported that installed base of smart assistant-compatible devices in the home, such as thermostats, light bulbs, and door locks, are projected to increase by 1.6 billion for the next three years. 

The growth of smart assistant-compatible devices, according to the report, is accredited to the voice control offered by the likes of Alexa and Google Assistant. Industry experts forecast an increase in the number of these devices in home automation and lighting (49 percent), home security and surveillance (18 percent), audio and video entertainment (13 percent), and others (20 percent).

“Growth will also be driven by the arrival of cheaper devices. Device sensors will give smart assistants new ‘senses’ that make them aware of the environment, leading to ever more use cases. For example, in lounges and kitchens, where more functionality will be unlocked by smart home devices working collectively with smart assistants,” senior analyst Jason Low said.

 

Trust and Privacy Concerns

Recently, a man named Shawn Kinnear from San Francisco claimed that his Echo activated by itself without him doing anything and that it cheerfully said: “Every time I close my eyes, all I see is people dying.” It was indeed creepy, and perhaps more worrying is that this isn’t an isolated case. A woman from Portland, Oregon found out that her Echo had taken upon itself to send recordings of private conversations to one of her husband’s employees. Amazon said in a statement that the voice assistant misheard instructions. 

These cases not only happen to ordinary people but also to companies. Over the past six months, Bloomberg, the Guardian, Vice News, and the Belgian news channel VRT have revealed that the “big five” tech companies—Amazon, Apple, Facebook, Google/Alphabet, and Microsoft—have been using human contractors to analyze a small percentage of voice-assistant recordings. While the recordings remain anonymous, they often contain information or details to identify or embarrass the user.

“We live in a techno-dystopia of our own making. If you still have an Alexa or any other voice assistant in your home, you were warned,” Matt Novak, a Gizmodo writer, declared. 

According to TechCrunch, an American online publisher focusing on the tech industry, a recent report from Microsoft showed that 41 percent of voice assistant users are concerned about trust, privacy, and passive listening. About 52 percent of those surveyed were worried that their personal information or data was not secure, while 24 percent said they don’t know how it’s being used. Another 36 percent said that they didn’t even want their personal information or data to be used at all.

41 percent of voice assistant users are concerned about trust, privacy, and passive listening / Photo by prykhodov via 123RF

 

The report showed that there’s a huge awareness around the potential violations and abuse of trust that could happen with voice assistants. Many users have also expressed their concerns about the increasing number of robocalls and malicious calls. Google Duplex revealed that the voice assistant software has such advanced AI that people are not able to distinguish between a machine and a human making a call. These kinds of scenarios could lead to more robocalls successfully scamming people. Some people can steal personal information from others by tricking them into thinking the robocall they have is with someone they know.

According to Total Security Daily Advisor, a free website and daily newsletter that provides help to technology and physical security professionals to advance their careers, conversational AI may also lead to privacy risks. Most of the time, tech companies assure their users that they will only keep the voice recordings long enough to analyze and understand what is said. However, it is possible that the data get stored for longer than most people would prefer. Automated systems that record users’ voices without their consent are both a privacy concern and a potentially illegal act.

 

Voice Assistants Are Still Deeply Imperfect

Loup Venture, a research-driven venture capital firm based in Minneapolis and New York investing in frontier technology, conducted a test that will analyze voice assistants. Results showed that Google Assistant answered 88 percent of the queries correctly, followed by Siri at 75 percent, Alexa at 72 percent, and Cortana at 63 percent. While these scores are high, this doesn’t mean that they will deliver effectively in the long run. 

Missing one correct answer out of every five or even 1 out of 10 attempts is more than enough reason to question a voice assistant’s effectiveness. This is the same reason why Google Lens, Google Now, and any number of Samsung phone features for the past years have never really taken off. Users stop using them after a while because, even as tech companies continually work on these devices, they still fail just enough to make them unreliable.