India plans AI policing tool targeting migrants
Maharashtra government links language-based surveillance to BJP pledge against “Bangladeshis and Rohingya,” raising alarms over profiling and mass misidentification
NEW DELHI, India (MNTV) — India’s ruling Bharatiya Janata Party (BJP) government in Maharashtra is developing an artificial intelligence system that would allow police to identify and detain people it labels as “illegal” Bangladeshi and Rohingya migrants — a move civil rights advocates say amounts to technology-driven targeting of Muslim communities under the guise of law enforcement.
The proposed system, announced by Maharashtra Chief Minister Devendra Fadnavis, is being developed in partnership with Indian Institute of Technology Bombay and is intended for statewide use by police agencies. Officials say the tool relies on linguistic analysis, examining speech patterns, tone and word usage to flag suspected undocumented migrants.
The initiative has drawn sharp scrutiny because it explicitly aligns with BJP political rhetoric that has long portrayed Bengali-speaking Muslims as “illegal Bangladeshis,” a label frequently used in India to question citizenship and justify detention, deportation or disenfranchisement.
At a recent political event unveiling the ruling alliance’s manifesto, Fadnavis declared that Mumbai would be “freed of Bangladeshis and Rohingyas,” language rights groups say reflects a campaign of demographic targeting rather than neutral enforcement of immigration law.
According to officials, the AI system has so far demonstrated an accuracy rate of roughly 60% after several months of testing — a margin of error that legal analysts and technology researchers warn could place thousands of Indian citizens at risk, particularly in a country where linguistic and cultural boundaries do not align with national borders. Bengali is spoken by millions of Indian citizens, especially in eastern India, and closely resembles dialects spoken across the border in Bangladesh.
The state government has not explained how the technology will distinguish between citizens and migrants who share language, accent and cultural markers, nor has it outlined legal safeguards to prevent wrongful detention. Authorities have also declined to clarify how speech data will be collected, whether through interviews, mobile devices, recordings or surveillance, or what oversight mechanisms will govern its use.
Despite these gaps, senior BJP leaders have already signalled plans to operationalize the system at scale. Party leader Kirit Somaiya has claimed that more than 224,000 birth certificates issued to alleged Bangladeshi nationals have been cancelled in the state and announced an upcoming campaign to identify and deport undocumented migrants following local elections.
Chief minister Fadnavis has also confirmed that detention centers are being established to hold those identified as suspected migrants prior to deportation, with land reportedly allocated for the facilities.
Rights advocates warn that combining error-prone AI tools with detention infrastructure risks institutionalising collective punishment, particularly against Muslim and Bengali-speaking populations.
Legal analysts note that the Maharashtra project reflects a broader trend under BJP rule, where technology is increasingly deployed to advance ideological objectives — from citizenship verification drives to digital surveillance — often without transparency or independent accountability.
In this context, civil rights advocates and legal analysts say artificial intelligence is being used not to improve governance but to automate exclusion.