Year in Review: Internet Shutdowns, Digital Privacy & AI Surveillance
Hey, it’s Sam. Welcome to our final newsletter of the year!
Accurately summarizing our collective experience of 2023 is basically impossible. But I fear Google’s autocomplete suggestions may offer some valuable hints…
From the launch of GPT-4 in March to the EU’s AI Act, the breakthrough of generative AI has dominated headlines this year. In fact, its influence has been so pronounced both Collins and Cambridge Dictionary chose AI-related terms as their respective words of the year. And despite AI’s ever-increasing pervasiveness, its real-world impacts are only just beginning to make themselves felt.
Meanwhile, there’s been a relentless barrage of stories revealing new ways our privacy is being undermined online: from monitoring people via smart phone notifications and the misuse of spyware, to the continued impact of the unregulated data broker industry on society’s most vulnerable.
On top of that, people from Azerbaijan to Zimbabwe have experienced internet disruptions as the trend of intentional internet shutdowns shows no signs of slowing.
Looking back on a year as bewildering as 2023 is not easy, so I reached out to four experts for help: Maria Xynou, Felicia Anthonio, Valentin Weber, and Steven Feldstein.
I asked each of them to tell me about a story or trend they thought was important from the past 12 months and how it’s likely to affect things going forward. Below I’ve compiled some of the most important issues we discussed.
Thanks for reading our newsletter this year. If there’s something you think we should cover in 2024, please get in touch at sam@top10vpn.com
Another Year of Internet Shutdowns
There’s no escaping the impact conflict has had on the world over the past 12 months. And in almost every country affected, there’s been a “heavy emphasis on ICTs [information and communication technologies],” says Steven Feldstein, Senior Fellow at Carnegie Endowment.
From debates around satellite internet services and the continuation of social media surveillance to deliberate internet restrictions, new technologies are undermining privacy and freedom of expression at vital moments.
Felicia Anthonio, #KeepItOn Campaign Manager at Access Now, says we’ve seen “the weaponization of internet shutdowns in times of conflict” during 2023 — a trend that is unfortunately set to continue throughout 2024.
One of the most overlooked internet shutdowns of 2023 occurred in the Amhara region of Ethiopia, where the internet access was disrupted throughout much of the year. A report by Access Now published earlier this year provides invaluable insight into how the disruptions directly impact people in the region.
We also saw a number of internet disruptions during exams, including in Syria and Iraq. Needless to say, blocking an entire platform or disconnecting everyone due to exams is a disproportionate response! Despite this, Access Now have been documenting the practice since way back in 2017.
Worryingly, the number of countries deploying this tactic increased this year, with Kenya joining the list in November when authorities blocked access to Telegram during national exams.
Given Kenya’s lack of pervasive internet censorship elsewhere, the restrictions were pretty surprising. “The nationwide blocking of a popular instant messaging app during exams (similar to other countries which have long implemented more pervasive forms of internet censorship) is quite concerning for a country like Kenya,” says Maria Xynou, Internet Censorship Researcher and Community Lead at the Open Observatory of Network Interference (OONI).
The team at OONI have been busy this year recording and analyzing internet disruptions in Jordan, Senegal, Brazil, Ethiopia, Pakistan and Turkey, to name just a few. OONI’s website was even blocked in China, as recorded by their own tool.
Meanwhile, they’ve also updated their OONI Explorer tool, added a new Censorship Findings Page and are working on updating their OONI Run tool next year.
2024 is set be dominated by elections, with crucial votes scheduled in a huge number of countries. And while these threaten yet more internet shutdowns, they also provide a unique opportunity “to hold those in power accountable [and] to uphold people's rights,” Anthonio says.
Early next year, we’ll be publishing our annual Cost of Internet Shutdowns report, which we hope will be helpful to those working to bring an end to internet shutdowns around the world. If you’re interested in learning more about the topic, I also highly recommend the new courses from Advocacy Assembly, curated by Access Now, IODA and other civil society organizations.
New Digital Surveillance Methods
Beyond the world of internet disruptions, digital privacy infringements have become ever-more sophisticated during the past 12 months. There’s been more revelations about the misuse of spyware technology from Citizen Lab and Amnesty International and earlier this month, it was revealed that governments were monitoring people via smart phone notifications.
It’s emblematic of “how crafty governments have become to snoop in an opaque way,” says Valentin Weber, Senior Research Fellow at the German Council on Foreign Relations.
It’s yet to be seen whether the revelations will lead to protective measures and the introduction of more robust privacy safeguards. What’s almost certain, however, is the continued evolution of surveillance tools that are able to identify vulnerabilities in the software we use everyday to spy on our communications.
“I see law enforcement inventing new avenues for getting access to information in 2024 and beyond, in ways that are not visible to citizens, especially with the proliferation of IoT in our homes and cities,” Valentin Weber.
With the ever-increasing number of “smart” technologies being deployed in our homes and neighborhoods, the number of potential vantage points from which we can be monitored continues to increase. With that, defense mechanisms need to be developed in tandem with the changing threats to maintain our privacy.
To read more about the ways increasingly opaque surveillance methods threaten our privacy, read Weber’s article on China’s Smart Cities.
AI’s Privacy Problem
It’s impossible to deny the huge impact AI tools are having in everyday life. And yet, a few stories from the end of this year point to the significant privacy implications the technology poses — whether that’s revealing training data via a security flaw, harvesting our sensitive data via commercial relationships with other companies, or being used directly for surveillance purposes.
What’s clear is that much more work is needed in 2024 to uphold our right to information and advance our right to privacy in the face of the rapid deployment of AI, the continuation of repressive internet shutdowns, and ever-increasingly advanced surveillance tools.
Until then, though, I think it’s probably best to try to rest and recuperate over the holidays as much as possible.
Holiday Reading Recommendations
Yifan Yao, et al: A Survey on Large Language Model (LLM) Security and Privacy: The Good, the Bad, and the Ugly
This paper examines the dual role of Large Language Models (LLMs) like ChatGPT and Bard in both enhancing and posing risks to security and privacy. It categorizes their applications into beneficial ("The Good"), offensive ("The Bad"), and vulnerabilities ("The Ugly"), noting LLMs' potential in improving code and data security, while also acknowledging their use in user-level attacks. .
Maanak Gupta et al: From ChatGPT to ThreatGPT: Impact of Generative AI in Cybersecurity and Privacy
Like the above, this paper delves into the cybersecurity implications of Generative AI (GenAI) models and discusses their use in both defensive and offensive cybersecurity contexts. It explores vulnerabilities, potential attacks (such as Jailbreaks, prompt injection), and defensive measures, while also addressing the broader social, legal, and ethical aspects of these technologies.
Kristin Eichhorn & Eric Linhart: Election-related internet-shutdowns in autocracies and hybrid regimes
This article examines the dual nature of the internet in autocratic and hybrid regimes, serving both as a tool for liberation and digital authoritarianism. It analyzes the strategic use of internet shutdowns during national elections, showing how these shutdowns are not random but are driven by strategic decisions and the varying degrees of uncertainty associated with each election.
Top10VPN’s 2023 Research Highlights
VPN Demand Surges Around the World
We’ve continued to monitor VPN demand around the world this year, with significant spikes recorded in Turkey, Senegal, the U.S., and Iraq. The live document details how these increases are often responses to government-imposed internet restrictions, social media blocks, and other forms of digital censorship.
Our 2023 VPN Vulnerability Report reveals a projected 44% increase in VPN vulnerabilities compared to the last year. The report details more than 300 VPN vulnerabilities disclosed since 2021, with Cisco, Zyxel, and SoftEther being the most affected vendors.
Russian Internet Censorship Spending Since 2022
Russia has increased its internet censorship spending by 63% after its invasion of Ukraine, with the most significant increases occurring near Ukraine's border. Our report, based on an analysis of almost 4,000 official procurement documents, reveals that Roskomnadzor and its federal departments have notably escalated expenditure in response to heightened internet censorship efforts, especially in regions critical to Russia's information control and broader internet censorship strategies.
Data Privacy Investigation: Chinese Electric Vehicle Exports
Our investigation into the data privacy risks of Chinese electric vehicles (EVs) reveals that these smart vehicles and their associated mobile apps pose significant privacy concerns. The research highlights issues such as inadequate privacy policies, extensive data collection including location and vehicle usage, and potential data transfer to China, emphasizing the need for awareness and improved data privacy protections in this rapidly growing market.