Tech Companies & Proactive Measures

The European Union is considering legislation to prevent the dissemination of terrorist content online. Certain members of European Parliament (MEPs) have expressed concern over the inclusion of a clause that would require tech companies to deploy “proactive measures” in an effort to stop the spread of terrorist online. These MEPs—often in hand with tech companies and their lobbyists—argue that such measures violate free expression. What is frequently missing from this discussion, however, is the fact that tech companies already deploy several proactive measures to promote a user-friendly experience, ensure user safety and satisfaction with the site or platform, and/or comply with laws. Below is a compilation of tech companies’ existing proactive measures that review and scan user content—many of which users encounter and rely on every day.

Harmful Content

Child exploitation content filters: Major tech companies and organizations around the world rely on hashing technology, known as PhotoDNA, to filter online platforms and sites for child exploitation material at the point of upload. PhotoDNA is able to analyze a known image quickly and accurately and prevents the upload of that piece of content. (Sources: Dr. Hany Farid, Microsoft)

Spam filters: These filters scan messages and organize email according to specific criteria as determined by the tech company. Filters were originally developed to filter out spam content to place in a separate folder or block it altogether. Some free email domains like Gmail often filter messages more aggressively. (Sources: Return Path, Mailchimp)

Suspicious links: WhatsApp, which is owned by Facebook, automatically performs scans on messages with URLs to determine whether or not they appear suspicious. The feature marks suspected websites with a red warning label in an effort to prevent users from clicking and going to a malicious site. (Sources: WhatsApp, Android Authority)

Virus and malware: Popular email services like Gmail and Yahoo Mail regularly scan all attachments that users send and receive for viruses and other malware. The email service provider will either attempt to clean or remove these malicious components from infected attachments, or prevent users from sending such attachments altogether. (Sources: PC World, Google, Yahoo)

Non-Harmful Content

Advertising: Tech companies review user data and content in order to serve specific ads to specific audiences. Moreover, tech companies have filtered ads due to inappropriateness. For instance, Google does not permit ads that contain sexually explicit content or child abuse imagery. (Sources: Google, Google, Vox)

Calendar events and travel: In 2015, Google announced that Gmail will scan email messages for emails with flight, hotel, restaurant, or ticket information. In an apparent effort to help users save time, Gmail then automatically ads that information into Google Calendar. (Sources: Google, Venture Beat)

Email filters (e.g. Gmail’s categories and Microsoft Outlook’s Focused inbox): These filters scan messages and organize email according to specific criteria as determined by the tech company. As mentioned above, filters were originally developed to filter out spam content to place in a separate folder or block it altogether. However, some email services have begun using filters to categorize messages to help users easily identify certain emails as important or urgent, a newsletter, a promotion, etc. (Sources: Return Path, Mailchimp)

Photo organization: Having a camera on your cellphone has made it easier to photograph daily activities. To help customers cope with their large collection of photos, tech companies like Google and Apple offer easy ways for their users to organize photos. Their platforms can scan users’ photos and organize them by date and even by specific people, places, or pets automatically. (Sources: Verge, Apple)

Purchases: Google’s Gmail scans for receipts in order to assemble a user’s purchase history. Facebook also performs a similar function scanning messages in Facebook Messenger. This data is mined for the tech companies’ use, but the industry giants have claimed that the data is organized for user convenience (e.g. track purchases, renew a subscription, etc.). (Sources: Fast Company, Google)

Recommended content: Social media platforms, notably YouTube, work to recommend content to users to keep people engaged and on their sites for longer. In YouTube’s case, the company’s algorithms analyze users’ watch history to make recommendations and queue an “autoplay” list. Other companies like Facebook actively recommend new friends to connect with based on several factors including mutual friends and interests—all user-based data. (Sources: New York Times, Wired)

Suggested replies: These features review your email messages and suggest an appropriate reply. The aim is to save its users time by providing short responses to emails that can be sent in a few clicks. These suggested responses are generated by a computer algorithm and use natural language processing and machine learning technologies. (Sources: Google, Microsoft)

Daily Dose

Extremists: Their Words. Their Actions.

Fact:

On November 29, 2020, an assailant detonated an explosives-filled military vehicle on an Afghan army base, killing at least 31 and wounding 24. 

View Archive

CEP on Twitter