The Ministry of Information & Broadcasting, through BECIL, last month floated (and closed) bids for an agency to provide “solutions and services” related to “fact verification and disinformation detection on social media platforms”. It’s not clear what the rush was, but BECIL made the EOI document public on May 13, bids opened on the afternoon of the same day, and closed after a week on May 20. An “Expression of Interest” document precedes an actual tender, bidders may “express their interest” in the project, and may approach the government with a more detailed proposal.
The bidding firm has a wide mandate:
- “Rapid fact verification” of text-based content, image, videos; identify geolocation and verify visual content, and identify fake news, and analyses web data against historical content.
- Identify “suspicious profiles”, pages and uploads of videos, images with potential to incite violence, “key influencers” behind disinformation, and their location via geolocation
- Identify/detect and monitor: identify and analyse automated and bot behaviours, identify and monitor key trends and hashtags used in spreading disinformation, monitor the activities of disinformation uploaders across multiple social media platforms, detect and analyse coordinated disinformation campaigns
- Analyse “malicious uniform resource locators and quick response codes”, metadata of files, images and videos with fake content; timeline and geolocation of disinformaton; carry out link analysis between different entities spreading fake news; sentiment analysis for large scale disinformation; carry out AI-based data classification and clustering; source code analysis of suspicious files in content; domain reputation analysis
The agency will have to develop a mobile-based “automated response framework” for “rapid querying of facts and images in existing repository of verified claims using Artificial intelligence”. The solution should be able to curate and accept training data using AI and ML. The agency will develop a system wherein it will run a piece of misinformation against a database of verified claims.
- Create corrective response with image, correct facts, and analysis
- Create periodical reports on fake news and disinformation content
- Archive and store all data of fact checked content/ images/ videos / response / reports and other data by means of online/offline archival support. The archive set-up thus have back up storage / archival and having long-term retention.
A few reminders
1. This is the government’s eighth attempt to monitor social media: This is the Indian government’s eight attempt to explicitly and directly monitor social media. Only this time, it is being done under the garb of detecting fake news and disinformation.
As recently as April 2018, the I&B Ministry had sought bids for a 800-person network for granular social media monitoring, going down to district-level monitoring. As with this proposal, the previous 66-page proposal sought an agency to listen and respond to online trends on social media platforms including Facebook, Twitter, blogs, and email, and sought creation of analytics reports and predictive analysis. The then I&B minister Rajyavardhan Singh Rathore had said that the hub is meant to “facilitate information flow regarding its policies and programmes through social media platforms”.
Starting 2014, the government has attempted to hire such agencies; BECIL itself — which says it has developed expertise since its formation in 1995 in “executing various turnkey based IT projects especially in the areas of Social Media Monitoring, Response management and other related services” — has floated at least three such tenders:
- In December 2015: to monitor individual social media user/account
- In February 2016: Sought bids asking for a “listening tool”
- In July 2017: Invited bids to “track overall trends on various social media platforms”
2. The I&B Ministry withdrew its previous proposal for social media monitoring, after the Supreme Court likened it to surveillance state: In a challenge to the April 2018 proposal, the Supreme Court had said that India will be “moving towards a surveillance state” if the proposal goes through. Mahua Moitra’s petition had argued that the proposal has no legal basis, defeats the right to privacy, and robs an individual of their identity. The government eventually withdrew the proposal, and indicated that it would be reviewed.
Some more detail:
1. The case is disposed off with a recorded undertaking by the Goverment it is withdrawning the tender.
2. The government indicated it will internally study the matter.
3. Yes, risks of subsequent tenders and direct procurement with a change in specifications.
— Internet Freedom Foundation (IFF) (@internetfreedom) August 3, 2018
3. Fake News is a common, and successful trope for the Indian government to push regressive policies through [Nikhil adds]: A couple of years ago, after the Cambridge Analytica scandal, and a spate of mob lynchings triggered by “fake news”, the IT Minister Ravi Shankar Prasad began a relentless demand for traceability, asking platforms to enforce traceability mechanisms, even if they might break end-to-end encryption. IT Minister RS Prasad has said that “If they [social media platforms] do not take adequate and prompt action, then the law of abetment also applies to them” and has said that “traceability will be WhatsApp’s job“.
The “WhatsApp lynchings” eventually triggered the proposed amendment to the Intermediary Liability Rules, which exempt platforms from the liability of their users’ actions. While the government said that their objective was not to weaken encryption, they knew that this was exactly what such a demand for traceability would do. In this context, it’s easy to forget that this was essentially a reboot of a previous attempt, in 2015, by the Indian government to weaken encryption.
Now it seems, the fake news trope has been brought up to bring in social media surveillance.
A few concerns
This is surveillance, plain and simple: Identifying ‘suspicious’ accounts, monitoring people across social media platforms, pulling their location, among other things; then using this data to analyse trends, link analysis of different people spreading disinformation, is all granular tracking of citizens’ activity, without a clear legal basis. Given the government track record, including that of the I&B Ministry and BECIL, it is likely that a more detailed tender will come out at a later stage, which will be far more overreaching.
Why does the government need geolocation? The indicated “aim” of the proposal seems to be to dispense counter-information, but what can justify the need to know the location of a person who may be spreading fake news?