It’s been a hard couple of years for the internet. Between election interference via social media platforms, disinformation going viral and facilitating genocides, and most recently, the rampant spread of a conspiracy theories that 5G internet broadband caused the COVID-19 pandemic, optimism in the internet’s force for good seems shaky at best.
I was scrolling Twitter recently and saw a tweet that ironically, mournfully noted, “Remember when this was the platform that was going to save democracy?” I do indeed remember. Simpler times, my friends.
Internet platforms such as Google and Facebook have invested immense resources in countering the more nefarious consequences of internet-enabled bad actors, including significant investment in machine-learning capabilities to uproot disinformation and box out cyber criminals before real harm occurs.
These supply-side focused solutions are indeed powerful, scalable tools in the fight against digital harms. However, to date, there has been less effort to develop demand-side solutions at scale, such as meaningful, global digital literacy and regulatory protections for users.
What Are User Perceptions of Internet Safety?
In an effort to kickstart a conversation around demand-side needs, our Center for Digital Acceleration (CDA) recently conducted research in Ghana and India on internet users’ perceptions of trust and privacy on the web. We sought to understand how users themselves parsed their own trust in internet platforms, what tools and tactics they used to keep their information private and themselves safe online, and what knowledge they had about certain aspects of internet use, such as platforms sharing user data.
Ultimately, our goal is to understand how we can ensure the internet is trusted—as one young man we interviewed in India pointed out, “if they don’t trust [the internet], no one will use it.” Trust is the foundational element of meaningful digital inclusion, and meaningful digital inclusion should be the end goal we all seek.
A talk I recently gave at the Global Digital Development Forum can be found below; high-level findings from our research can be found below. We will be publishing our full report on this topic this fall (subscribe here to follow updates).
High-level Findings from Ghana
The internet is generally trusted: Most focus group participants felt they had absolute control over social media and email accounts due to the use of passwords. This perception is widespread thus the internet is used without hesitation.
Peer effect of trust: If apps were popular and many of their friends used or trusted them, they felt comfortable trusting them, too. A clear example is Whatsapp.
Low awareness of data sharing: There was very low awareness of platforms such as Google and Facebook sharing user data with third-party advertisers. When prompted, focus group participants felt insecure and disappointed and that their privacy terms had been breached.
Control and agency: Many participants felt they had full agency over their social media accounts because they decided what content they shared and how they engaged in that space; additionally, the use of passwords to access accounts gave them a real sense of control.
High-level Findings from India
Privacy is utopian…but the internet is fun: When asked about what they think about privacy on the internet, many respondents expressed feeling like it was a utopian dream. In other words, noting that it was unlikely. However, they all continued to engage online as it’s fun and social and offers value to their lives.
Trust by feature: Interestingly, Indian focus group participants based their trust of a platform both off of word-of-mouth popularity as well as by specific privacy features, such as end-to-end encryption.
Passwords and settings: Because phone sharing was also an aspect of the population we sampled with our focus group, there was an advanced understanding of how to navigate a phone’s more complex password and privacy settings, from erasing browser history to using app-specific passwords and changing out SIM cards and contact data.
This post originally appeared on ICTWorks on July 1, 2020.