Social and Ethical Concerns of Smart Voice-Enabled Wireless Speakers


The following article explores the technological convergence of Intelligent Personal Assistant (IPA). More specifically, it analyzes the emerging technology of Smart Voice-Enabled Wireless Speakers and how the IPA technology transitioned from smartphones to built-in, dedicated home devices. The trend is changing the way in which machine learning and artificial intelligence are used, and the article explores the new and evolving technology in the context of its overall social impact.

Intelligent Personal Assistants (IPA)

Before we explore Smart Speakers in the context of their social and ethical impact, let’s first cover the basic introduction to Intelligent Personal Assistants.

Today’s intelligent personal assistants typically come in the form of software installed on a smartphone (Apple Siri, Google Assistant, Samsung Bixby) or embedded as part of the operating system (Windows Cortana). The role of an intelligent personal assistant is to essentially act as a service that listens to user voice or written questions and commands. The user input is collected and parsed by the underlying software, which is connected to various online application program interfaces (such as weather, news, traffic, sport, stock or imagine and shopping APIs) and is typically also using the user location to provide the better service.

The main challenge of the IPA solutions remains in the area of incorrectly apprehending the language and parsing the voice input. Another challenge is in the area of understanding the context and the overall meaning of the user request, without which it is not possible to perform the desired action.

Regrettably, the current AI algorithms are still not very performant at making sense in the huge variety and volume of the information. A terrible example of this is a recent case where Google Assistant “applied automatic labels to pictures in digital photo albums, classifying images of black people as gorillas.” (Crawford, K., 2016). While Google expressed regrets, explaining the random nature of the issue, errors such as these tend to undermine the public trust.

However, the worldwide smartphone market keeps growing at an unprecedented pace. In 2017 alone, the shipments for smartphones equipped with IPA are predicted to hit an all-time high of over 300 million devices (Intelligence, B., 2017), and the smartphone marketplace anticipates over 6 billion smartphones to hit the market between 2018 and 2021, with the majority of them pre-installed with IPA.

Fortunately, the problem of smart technology intelligence is somewhat alleviated by the emergence of Big Data and Artificial Intelligence technologies, which apply machine learning algorithms to increase the reliability of IPA services. As a matter of fact, these technologies are gaining a tremendous traction and is especially applicable to the field of Intelligent Personal Assistants, where costumers expect “improved ability to anticipate their needs and enhance applications such as indoor navigation, speech recognition, and language translation, even when not connected to network” (Neal P, Fenech C., 2017).

Examining the customer demands for better and smarter technology solutions, we can assume the further proliferation of devices equipped with onboard AI and neural machine learning capabilities.

Smart Voice-Enabled Wireless Speakers

The Smart Voice-Enabled Wireless Speakers are an entirely new branch of devices representing the convergence of technology from smartphones and operating systems, into the whole new realm of dedicated AI devices. The Smart Voice-Enabled Wireless Speakers are the newest and most creative movement among IoT gadgets presently entering the marketplace.

The current market of smart speakers is represented by the devices such as Amazon Echo, Google Home and Lenovo Smart Assistant (Figure 1), but also by the lesser known Mattel Aristotle, LG Hub Robot or upcoming Microsoft Voice-Activated Cortana speaker.

Figure 1 – Martin, (2017)

The market research firm Gartner estimates that by 2020, global sales of smart wireless speakers “that listen and respond to voice commands, will reach approximately 2$ billion dollars and over 25% of all households will use two or more of these devices“ (, 2017).

What is the premise of a Smart Speaker?

Well, in short, the speaker initializes whenever it hears the wake word, such as “Echo” in the case of Amazon Echo device or “OK Google” for Google Home smart speaker. From there it is up to the user. All that needs to happen is to tell the device what you need. “Whether that is some light jazz, the latest headlines from NPR, a 20-minute kitchen timer, a particularly dumb joke or any one of the countless other things you might think to ask for, it is entirely up to you.” (Crist, R. and Carnoy, D., 2017).

While the smart speakers are effectively still powered by the same software used in IPA mobile applications (only slightly tweaked for the home use) the most significant is the convergence of form. Smart Speakers are perhaps the first devices ever, where we can see a transformation from a mobile application to a standalone dedicated appliance, that is entering our homes, branded as a futuristic Artificial Intelligence IoT device.

Social and Ethical Impact & Concerns

Any new and significant technological progress has a tendency to introduce various concerns.

The concerns for individual freedoms are even more intensified and especially hold true with devices such as smart speakers, equipped with microphones that are capable of constantly listening and recording every single word spoken in our households.

Aside from autonomy and privacy concerns of the smart speaker devices, which are obvious to most of us, the following are some additional, less often mentioned additional concerns that I see as valid issues that also carry social and ethical implications, that can affect individuals or more broadly entire society.

Data Ownership

An excellent example of a data ownership concern is an ongoing court case in Arkansas, where authorities are demanding the release of all recordings from Amazon Echo device, to investigate a murder case (Eliott C. McLaughlin and Keith Allen, C., 2017). In the above case, prosecutors were able to secure the warrant for Amazon Echo data, which is asking for a copy of all recording that happened at the residence in question. “A search warrant states that police believe they could have extracted audio recordings, transcribed records, text records, and other data.“ (The Verge., 2016).

The question that immediately comes to mind is that of data ownership concern. Are we, as a society comfortable with a home assistant that can testify against us in court?

Data Security

Most smartphone devices do not come equipped with the extended storage necessary to hold all the audio recordings. That is why smart speaker devices typically run in the “cloud,” and neither is the processing of commands completed locally on the device itself, meaning that most of these devices are unusable without an internet connection. That said, the most significant concern when it comes to data security is that of all home audio recordings stored on the device or our premises.

Security of cloud-based applications and data is one of the key concerns of cloud customers (Sathyanarayana, T. V., & Sheela, L. M. I., 2013, December).

What if there’s a data breach and our home audio recordings are exposed publicly?

Cyber Bullying, Cyber-harassment & Ransomware

When referring to data ownership and security, I believe that smart speakers will be especially susceptible to Cyber-harassment and Cyber-Bullying as a possible form of hacking.

A hacked smart speaker can record every word we say; and become compromised without our knowledge. Regarding smart speakers, we do not have the option to install additional antivirus and firewall services to protect our privacy further, so the probability of an intruder hacking the smart speaker device is higher than that of a mobile phone.

With the smart speaker connected to the internet, recording all private conversations, it is possible that compromised audio content could be used as a form of cyber bullying.

Nevertheless, the smart speaker device could also become a perfect ground for a Ransomware malware attacks, through which intruder would be able to extort money for releasing or erasing user audio recordings in question.




Academic References

Incerti, F. (2017). Amazon Echo: Emerging technology for formal or informal learning?. Association for the Advancement of Computing in Education (AACE), [online] pp.1627-1633. Available at: [Accessed 27 Mar. 2017].

Sathyanarayana, T. V., & Sheela, L. M. I. (2013, December). Data security in cloud computing. In Green Computing, Communication and Conservation of Energy (ICGCE), 2013 International Conference on (pp. 822-827). IEEE. [Accessed 27 Mar. 2017].

Erdur-Baker, Ö. (2010). Cyberbullying and its correlation to traditional bullying, gender and frequent and risky usage of internet-mediated communication tools. New media & society, 12(1), 109-125. [Accessed 28 Mar. 2017].

Crawford, K. (2016). Artificial intelligence’s white guy problem. [Accessed 28 Mar. 2017].

Other References

Neal P, Fenech C. (2017). Digital Predictions 2017 – The Deloitte Consumer Review – New Technology – UK. [online] Available at: [Accessed 28 Mar. 2017].

Intelligence, B. (2017). THE GLOBAL SMARTPHONE REPORT: The forces behind the global deceleration in smartphone sales. [online] Business Insider. Available at: [Accessed 29 Mar. 2017].

Martin, (2017). Here are the best alternatives to the Amazon Echo which are coming in 2017. [online] PC Advisor. Available at: [Accessed 30 Mar. 2017].

Crist, R. and Carnoy, D. (2017). Amazon Echo review: The connected home experience you didn’t know you wanted. [online] CNET. Available at: [Accessed 30 Mar. 2017]. (2017). Gartner Says Worldwide Spending on VPA-Enabled Wireless Speakers Will Top $2 Billion by 2020. [online] Available at: [Accessed 30 Mar. 2017].

Eliott C. McLaughlin and Keith Allen, C. (2017). Alexa, can you help with this murder case?. [online] CNN. Available at: [Accessed 30 Mar. 2017].

The Verge. (2016). Police want an Echo’s data to prove a murder case, but how much does it really know?. [online] Available at: [Accessed 30 Mar. 2017].