Token Disclosures?
Token Disclosures? is a The Hindu analysis published on 4 August 2013 by Deepa Kurup. The article examines how Edward Snowden’s revelations about the NSA’s Xkeyscore surveillance programme exposed the limitations of corporate transparency reports, suggesting that government access to data through direct surveillance infrastructure rendered platform disclosures incomplete.
Contents
Article Details
- 📰 Published in:
- The Hindu
- 📅 Date:
- 4 August 2013
- 👤 Authors:
- Deepa Kurup
- 📍 Location:
- Bangalore
- 📄 Type:
- Analysis
- 📰 Newspaper Link:
- Read Online (Subscription required)
Full Text
Snowden's Xkeyscore expose makes a mockery of Twitter's transparency revelations
This week, roughly around the same time, two 'revelations' made headlines in the world of technology. The first, the U.S. National Security Agency's top secret web surveillance programme, codenamed Xkeyscore, another expose from the house of Edward Snowden & Co.; and second, microblogging site Twitter's third biannual Transparency Report for the first half of 2013.
The former exposed a global surveillance net, cast far and wide to freely (no formal authorisation required) access and mine emails, chats and browsing histories of millions. The content of the latter report not only pales in comparison but also raises fundamental questions on just how much goes on beyond the arguably modest claims made on Twitter's transparency charts.
Documents published by The Guardian have the NSA claiming that the "widest-reaching" system mining intelligence from the web had, over a month in 2012, retrieved and stored no less than 41 billion records on its Xkeyscore servers. These mind-boggling numbers make a mockery of Twitter's few hundred access request disclosures, advocates of online privacy and freedom point out. Then, it is hardly surprising that a large chunk of global requests came from the U.S. government: no less than 902 of the total 1,157 requests, accounting for 78 per cent. A far second is Japan at 8 per cent followed by the U.K.
India references
Interestingly, both Twitter's report and the NSA's Xkeyscore document have India references. While a map titled 'Where is Xkeyscore' in the training manual released showing India as one of 150 sites (hosting a total of 700 servers) indicates that India's very much on the global surveillance radar of the United States government; the fact that the India is a new entrant on Twitter's 'Country Withheld Content Tool' means that the government here is also making active interventions in microblogging content. This is very much in line with stances the Indian government has taken over the last year, swinging indecisively between asking internet firms to pre-screen content and asking service providers to take down what it finds offensive.
India, a bit-player
The Twitter report states that over the last six months it has seen an increase in the number of requests received (and eventual withholding of content) in five new countries: India, Brazil, Japan, Netherlands and Russia. In terms of numbers, India is still very much a bit player in the game given it falls under the 'less than 10 category, a list where the number of requests for user information made by the government during this period is fewer than 10. It appears from the report that Twitter did not honour any of these requests, indicating that either the requests were too broad or failed to identify individual accounts.
In the same period, Twitter received two requests from India to remove content, one from the "government/law enforcement agency" and the other through a court order. In all, three tweets were removed by Twitter. No details on the nature of content removed were available.
Transparency trends
A late entrant to transparency initiatives, Twitter's bi-annual reports have been applauded by privacy activists as an initiative that at least attempted to offer a glimpse into the otherwise opaque medium/industry. According to 'Who Has Your Back' an initiative by the Electronic Frontier Foundation, which tracks which corporate helps protect your data from the government, only a third of the 18 internet majors publish Transparency Reports – in fact, Facebook, WordPress and Tumblr all don't publish.
While it's definitely good that Twitter's providing data for India, post-Edward Snowden and his revealing PRISM leaks, netizens would question to what extent this data is representative of the magnitude or extent of user data tracking. Do governments like the U.S. need to approach Twitter (or other internet service providers) at all to access detailed user activity logs, content and metadata?
Secret orders excluded
Twitter makes it clear that its current report does not include "secret orders" or FISA disclosures. In another blog related to the Transparency Report, Jeremy Kessel, Manager, Legal Policy at Twitter Inc, writes that since 2012, Twitter's seen an uptick in requests to withhold content from two to seven countries. He writes that while Twitter wants to publish "numbers of national security requests – including FISA (Foreign Intelligence Surveillance Act) disclosures – separately from non-secret requests." It claims it has "insisted" that the United States government allow for increased transparency into "secret orders". "We believe it's important to be able to publish numbers of national security requests – including FISA disclosures – separately from non-secret requests." Unfortunately, we are still not able to include such metrics, Twitter states.
'Not the whole truth'
In the absence of these metrics, Sunil Abraham, director of Centre for Internet and Society, feels transparency reports "may not tell us the whole truth". The Xkeyscore revelations then may explain why the U.S. government has made only 902 information requests. "A programme like Xkeyscore potentially allows them to capture the very same data without having to approach Twitter. This is the very same imperative behind the CMS project in India. Governments across the world want to automate private sector involvement in blanket surveillance measures so that it won't serve as a check on their unbridled appetite for data"
He warns that there's a likely "race to the bottom", given that an unintended consequence of transparency may be that governments, rather than being shamed into respect for free speech and privacy, would be emboldened by the scale of surveillance and censorship in the so-called democracies such as the US and EU members that are on top of the global blanket surveillance game.
Context and Background
The article appeared during the peak of Edward Snowden’s NSA revelations in summer 2013. Following the initial PRISM disclosures in June, The Guardian published classified documents in late July detailing Xkeyscore, described as the NSA’s “widest-reaching” surveillance system. The documents revealed that Xkeyscore had collected 41 billion records in a single month during 2012, enabling analysts to search emails, chats and browsing histories without prior authorization.
Twitter had released its third biannual transparency report for the first half of 2013 concurrently with the Xkeyscore revelations. The report disclosed 1,157 government requests for user information globally, with 902 (78%) originating from the United States. However, Twitter explicitly noted that the report excluded Foreign Intelligence Surveillance Act requests and other classified legal demands, which the company was prohibited from disclosing under US law.
This exclusion fundamentally undermined the credibility of transparency reporting as an accountability mechanism. If intelligence agencies could access user data directly through infrastructure-level surveillance programmes like Xkeyscore and PRISM, the volume of formal requests submitted to platforms became meaningless as a metric. India appeared in both contexts: the Xkeyscore training manual identified India as hosting surveillance infrastructure sites, whilst Twitter’s report showed India making its first content removal requests.
Sunil Abraham’s analysis connected these revelations to India’s Central Monitoring System project, suggesting a global trend toward automating surveillance without requiring individualized legal process or platform cooperation, thereby circumventing potential corporate resistance as a check on government data collection.
External Link
- Read on The Hindu (Subscription required)
📄 This page was created on 20 January 2026. You can view its history on GitHub, preview the fileTip: Press Alt+Shift+G, or inspect the .