Threat modeling - what people are concerned about and best practices for mitigation?


Encryption illegal Corporate agents (ISPs) UAE political influence (blaspmy content filtering on FB, twitter, etc.) Tech dissidents are recognized (assumed) and potentially targeted by ISI

Not just for themselves, but as a network entry point for other potentially "problematic"

Tor not specifically prohibited Higher/targeted threat is hidden, abitrary, and assumed: people disappeared, tortured, used as a deterrent/example against organized activism Perpetual state of emergency Case examples

People targeted for liking comments on facebook Comparative religion courses banned If you are a journalist (subset: targeted group), Tor, etc., can be protective

Military balance of power precarious (power/popularity depends on social

Activists fight both (civil) goverment and military threats

Privacy in constitution, but excepted for "national security" No data protection law or privacy law in penal code Whats App technically illegal because of

Monitoring and Reconciliation of Telephony Traffic Regulations, 2010

Data is inaccessible, therefore tool of encryption is prohibited

1996 Fair Trial Act

Motivation: Pakistan (state) sees itself as (and can act as if) global defender of Islam NGOs suspect: must be geotagged


Similar to pakistan: fascism, encryption (and bollywood) Not as extreme as Pakistan Internet shutdowns: Marginalization, isolation Usage makes you a target (possible because you stand out) Few security/privacy activists

If they disappear, both event and impact may not be recognized

Targeting is also assumed arbitrary

"Maoism" is an excuse for suppression

Terrorism/state security is an excuse for targeting activists

Technology activists currently not especially targeted, but

No expectation of right to privacy (officially enforced) Very tech savvy: India-specific communications, encyption, qr

Allegedly: India tech is more secure than global tech

But no accountability, answerability, transparency

High-level encryption must be officially approved and should be assumed compromised at state level Low-level encryption permitted

Other threats/examples

For Iran, Ethiopia, Tor is explicitly prohibited Uzbekistan: Global certificate Golden (encryption) keys, back doors UK data spying acts Hardening of borders Legal requirement to use specific technologies

'90s South Korea: Active X as SSL Brazil: IE

Bubbles/echo chambers: how to encourage diversity of voices/views?

Balancing factors

Global corporate competitiveness Enforceability: Encryption must be arbitrarily enforced, because it's necessary for everyday security. Typically compliance is used as a lever: if gov't wants data, law is invoked.


Pseudonyms (however, name recognition is important for credibility, recognition, and gravitas) Mesh networks (e.g., Fire Chat in Hong Kong protests) as alternate social media mode (however, temporary solution, often not scalable, may be unstable) Tool development subject to funder priorities (which may not meet all risk-abatement goals) Avoid network fingerprinting by setting up a bridge (esp. private bridge) Compartmentalization (what you do outside of threat environment vs inside of it) Translation into the people's languages

Pro example: TAILS documentation and web site

UX: hierarchical information. Core pages: If you're going to translate something, make it this.

Outreach, communication skills, and media/PR competency Strong/clear/resonant arguments for privacy

Good for your kids India Net Neutrality example (ALB) John Oliver (Net Neutrality, Gov't Surveillance)

UK: "I never thought a leopard would eat my face cries woman who voted for leopard-face-eating-party."

Draw people into healthier social media apps by features other than privacy (e.g., emojis, not privacy)

Diversity of social media platforms


Localized (language) digital privacy/security training materials and documentation

Images Video

Better messaging/arguments for privacy

Last modified 2 years ago Last modified on Mar 27, 2017, 1:01:28 PM