censorship-course

Internet Censorship Course

View the Project on GitHub noise-lab/censorship-course

BGP and Web Manipulation

Format. Break into groups of 4–5. Each group picks one of the two breakouts below (or takes both if time allows). Spend ~5 minutes skimming the prep reads, then ~10 minutes debating. A designated reporter brings the group’s position — and any dissents — back to the full class for a ~3-minute report-back.

The book’s BGP section is a parade of accidents that look like attacks (AS 7007, Pakistan Telecom, China Telecom, Facebook 2021) and attacks that look like accidents. The web-manipulation section is quieter but more intimate: block pages tell you exactly who is censoring you. These breakouts push on the two extremes — the coarsest tool (BGP) and the most user-visible one (block pages) — and ask what obligations flow from operating the pipes.


Breakout A: Should Governments Mandate RPKI?

Motion. “National regulators should legally require every ISP operating in their jurisdiction to deploy RPKI-based route origin validation, with penalties for non-compliance.”

Prep reads (5–10 min).

Discussion prompts.

Bring back. Whether your group would support such a mandate in your own country, and what one carve-out or safeguard is non-negotiable.


Breakout B: Kill-Switch Orders and ISP Refusal

Motion. “ISPs served with a government order to withdraw BGP routes or block major services should refuse — even at the cost of their operating license.”

Prep reads (5–10 min).

Discussion prompts.

Bring back. The circumstances (if any) under which your group thinks an ISP is obligated to refuse a shutdown order, and how you’d operationalize that.


Instructor notes

These map to §2.2.3’s “all-or-nothing” characterization of BGP as a censorship tool and the Egypt/Pakistan case studies, plus the §2.2.4 point that block pages make censorship user-visible. Breakout A is the more technical debate; Breakout B is the more moral one. If the class has been engaging well with the “who gets to pull the lever” thread from Lecture 1, B lands with more force. If the class is technically inclined and has just read the RPKI/MANRS material, A is the better bet. Either way, the Pakistan Telecom hack (advertising /25s to reclaim traffic) is worth putting on the board — it’s a great concrete illustration of how ad hoc BGP defense really is.