Surveillance Capitalism: How Big Tech Owns You | Omar Al Zabir

Summary

The discussion examines how large technology platforms collect and use personal data, the security limits of phones and messaging apps, social-media harms, and practical ways to reduce exposure. Omar Al Zabir draws on his former work on Meta’s high-severity harms team and his current work at Kahf. Many of the statistics and security claims below are his, as presented in the interview.

How digital surveillance works

• Al Zabir describes Google, Apple, and Meta as operating an extensive surveillance system built from activity across apps and websites. He says collected data can include location, searches, viewing and scrolling habits, purchases, and interactions with content.
• He cites the presence of Google and Meta trackers on many websites and apps, and argues that data from sources such as maps, weather, and prayer apps can be combined to build detailed profiles.
• In his account, AI systems can use these profiles to infer people’s interests, moods, and likely behavior, then tailor content or advertising accordingly.
• He warns that targeted feeds can encourage compulsive use, reinforce ideological “bubbles,” promote harmful material to vulnerable users, and influence what people feel comfortable saying online. He also notes that algorithmic decisions can affect access to ads and opportunities.
• He says governments can seek data through legal processes such as warrants and emergency requests. He also discusses geofence warrants and cell-site simulators as ways authorities may identify devices in a particular area, including at protests. His concern is that association with a person or group can expose otherwise uninvolved people to scrutiny.

Phones, apps, and messaging security

• Separate sensitive and everyday use: Al Zabir recommends keeping sensitive apps and data separate from less-trusted apps, potentially on different phones. His reasoning is that a vulnerable app could provide a route to compromise the wider device.
• WhatsApp encryption versus device security: He says WhatsApp’s end-to-end encryption is strong, but encryption cannot protect messages if a phone itself is compromised. He describes several other risks: traffic patterns that may reveal group membership, people joining groups to collect member lists, spyware, and forensic access to a device.
• Signal: He prefers Signal on principle and says it stores less data than WhatsApp, but stresses that it does not prevent device compromise. He recommends disabling message previews in notifications because previews may remain cached outside the messaging app, even after a message is deleted.
• Operating systems: He disputes the idea that iPhones are invulnerable, citing zero-day exploits and spyware as examples of risks to any device. For high-sensitivity use, he recommends GrapheneOS, while acknowledging that security depends on more than the operating system alone.
• Smart TVs: Al Zabir warns that smart TVs may collect viewing or audio data. He cites a US FTC case involving Vizio tracking viewing activity and refers to claims in the Vault 7 leak about Samsung TV microphones. These surveillance claims are presented by him in the interview.

VPNs, Tor, and browsing

• Al Zabir says a VPN mainly hides a user’s IP address and location from some services; it does not necessarily prevent tracking, and the VPN provider may have browsing logs. He expresses particular distrust of many VPN providers and names Proton VPN as one he considers more privacy-conscious.
• For sensitive browsing, he recommends Tor, which routes traffic through multiple relays. He cautions against using Tor while simultaneously logged into identifiable accounts or browsing normally in Chrome or Safari, since activity may be correlated.
• For more sensitive research, he suggests Tails OS, a USB-booted operating system designed to leave little or no activity on the computer after shutdown. He warns that logging into personal accounts while using it defeats the anonymity goal.
• For routine use, his simpler recommendation is a privacy-focused browser rather than Chrome, Safari, or Edge. He mentions Brave and DuckDuckGo and says Kahf has also modified a browser to remove Google tracking and telemetry.

Social-media design and ways to reduce use

Al Zabir links social-media harms to business incentives: platforms earn advertising revenue by attracting users and keeping them engaged. He describes features such as infinite scroll, autoplay, and carefully sequenced content as mechanisms that can encourage continued scrolling and reinforce dopamine-driven habits.

His suggested personal steps include:

• Set daily app limits, particularly for social media.
• Turn off autoplay and sound.
• Switch the phone display to grayscale to make apps less visually enticing.
• Use social-media websites instead of their apps, since the less polished experience can add friction.
• Consider what a feed is showing—and what it may be omitting—rather than treating it as a neutral or universal selection.

Children’s online safety

Drawing on his Meta experience, Al Zabir describes exposure to grooming, sexual exploitation, harmful content, and other serious risks. He cites a study reporting more than 300 million child victims of online sexual abuse globally, and says Meta detected 6.6 million instances of child sexual abuse material in the EU in 2022. These figures are cited by him during the interview.

His main practical advice for parents is:

• Do not give children unsupervised internet-connected devices in their bedrooms, especially at night.
• Keep computers, tablets, and phones in shared spaces, with screens visible to adults.
• Be cautious about chat apps that let children communicate privately with strangers; he specifically warns against Snapchat.
• Delay children’s access to social media. He says he would not recommend it for children below about 15 or 16, while acknowledging that families and circumstances differ.
• Learn to recognize early signs of distress and seek guidance on online safety.

AI chatbots and deepfakes

• Sensitive information: Al Zabir says major AI providers may meet stringent security and compliance standards, but advises against entering sensitive personal information because data can still be exposed, misused, or accessed by insiders. He recounts an incident in which a token used with a Chinese AI model was exposed and attackers accessed servers.
• Deepfakes: He argues that image and video generators are advancing faster than detection tools, making realistic fakes increasingly difficult to identify. He also points to the difficulty of determining consent and the spread of deepfake material in encrypted or otherwise hard-to-monitor channels.
• AI companions and children: He describes cases involving chatbot platforms and young people, and warns that automated systems could help groomers contact many children and identify vulnerable targets at scale.

Kahf products discussed

Al Zabir presents Kahf as a company developing digital-safety products for Muslim and non-Muslim families:

• Kahf Guard: A mobile app that the company says blocks access to a database of 7.3 million harmful websites, blocks trackers and ads, and lets users restrict app use. Al Zabir says it has more than 3.5 million downloads.
• Kahf Browser: A browser intended to combine privacy, filtering, and content controls. He says it removes Google tracking and can blur indecent images or video as they appear.
• Kahf Kids: A child-focused video app featuring a curated collection of YouTube videos. Al Zabir says the team reviewed more than 100,000 videos and that the app removes YouTube advertising.
• Router and devices in development: At the time of the interview, Kahf was developing a home router intended to filter harmful sites and enforce internet or social-media time limits. A child-focused tablet and phone were also described as planned products.

Practical guide and recommendations

1. Separate sensitive activity from general app use.
2. Disable message previews in notifications.
3. Use a privacy-conscious browser and avoid unnecessary social-media apps.
4. Treat VPNs as limited privacy tools, not complete anonymity. Use Tor or Tails only when appropriate, and avoid identifying logins.
5. Limit social-media time and add friction through grayscale, disabled autoplay, and website use.
6. Keep children’s internet-connected devices in shared spaces and out of bedrooms at night.
7. Avoid sharing sensitive personal information with AI chatbots.

Main speakers and sources

• Omar Al Zabir — Guest; former Meta engineering manager on the high-severity harms team and co-founder and CTO of Kahf.
• Muhammad Jalal — Host and interviewer.
• Sources and examples cited by Al Zabir include Shoshana Zuboff’s book The Age of Surveillance Capitalism; reports or findings he attributes to Amnesty International, the US FTC, UNSW Sydney, the BBC, and the UK Home Office/National Crime Agency; as well as Meta safety work and his own professional experience.