A new study examining internet censorship practices in India has revealed significant inconsistencies in how websites are blocked across different internet service providers (ISPs). The research suggests that millions of users may experience different levels of access to the web depending on the network they use.

The report, titled “Poisoned Wells: Examining the Scale of DNS Censorship in India,” was authored by Karan Saini and supported by the Open Technology Fund under the Information Controls Research Program hosted at the Georgia Institute of Technology.

Over 43,000 Domains Found Blocked

The research confirmed that 43,083 unique domains are currently blocked through DNS-based filtering in India. This figure is significantly higher than earlier estimates, representing a sixfold increase compared with the previous largest study that documented 6,787 blocked domains.

To reach this conclusion, the researcher conducted an extensive technical investigation:

• Over 230 million domain names were tested

• 1.76 billion DNS queries were executed

• Tests were conducted across six major Indian ISPs and telecom operators

• The analysis was carried out between March and June 2025

The results provide one of the most comprehensive insights into website blocking practices in India to date.

DNS Filtering: The Primary Blocking Method

The study found that most website censorship in India relies on Domain Name System (DNS) filtering.

DNS acts like the internet’s directory system, translating domain names into IP addresses that allow devices to access websites. When DNS filtering is applied, the ISP’s server returns incorrect or empty results when users attempt to access blocked domains.

While relatively simple to implement, DNS filtering can be inconsistent and easier to bypass compared with more advanced methods such as server name inspection or deep packet filtering.

According to the report, India’s legal framework does not mandate a specific technical method for implementing website blocking. As a result, ISPs adopt different filtering techniques depending on their internal capabilities and operational decisions.

This flexibility has led to inconsistent censorship experiences for users across different networks.

Cases of Over-Blocking

The study also identified several cases where filtering appeared to be broader than intended.

For example:

• MTNL, now merged with BSNL, was found to block services such as Telegram, even though the platform remains accessible on other networks.

• The report suggests that web filtering lists intended for government offices may have been applied to residential users in some cases.

Another example involved Bharti Airtel, where the entire .yokohama top-level domain was reportedly blocked. This meant all websites using that extension became inaccessible to Airtel users.

Researchers indicated that this may have resulted from a technical misconfiguration rather than a deliberate policy decision.

What the Findings Suggest

The report highlights how decentralized implementation of censorship policies can produce uneven results across networks.

Key implications include:

• Users on different ISPs may experience different versions of the internet

• Technical errors can lead to large-scale over-blocking

• Lack of transparency makes it difficult to understand why certain sites are inaccessible

Since blocking orders in India are typically confidential and ISPs are not required to publicly disclose restriction lists, measuring the full scope of website censorship remains challenging.

A Growing Area of Debate

The findings contribute to a broader debate about internet governance, transparency, and consistency in digital regulation.

As India continues to expand its digital infrastructure and internet user base, questions about how website blocking is implemented—and how accurately it is enforced—are likely to attract increasing scrutiny from researchers, policymakers, and the technology community.

The study suggests that clearer technical standards and transparency measures could help reduce unintended over-blocking while ensuring that lawful restrictions are applied consistently across networks.