The Maharashtra government, in partnership with IIT Bombay, is developing an AI-based language verification tool to help identify suspected illegal Bangladeshi nationals and Rohingyas, a move that has sparked debate over accuracy, ethics, and civil liberties.
The Maharashtra government has announced the development of an artificial intelligence (AI) tool that will analyse speech patterns, tone, accent, and linguistic usage to help identify suspected illegal Bangladeshi nationals and Rohingyas living in the state.
The project is being developed in collaboration with the Indian Institute of Technology (IIT) Bombay and is intended to assist law-enforcement agencies during preliminary verification processes.
Chief Minister Devendra Fadnavis has described the initiative as a technological aid to address concerns over illegal immigration, which he has repeatedly flagged as a security and administrative challenge for the state.
According to officials familiar with the project, the AI system will not be used as standalone proof of nationality but as an initial screening mechanism to flag cases requiring further investigation through established legal procedures.
The project is currently at an experimental stage and has not yet been deployed on the ground. Officials say it will be refined over the coming months before any wider rollout is considered.
How the Language-Based AI Tool Works
According to senior officials in the state’s information technology department, the AI tool is being trained to detect linguistic markers such as accent, pronunciation, sentence construction, word choice, and tonal patterns.
These markers are then compared with known linguistic datasets to assess whether a person’s speech aligns more closely with certain regions outside India.
The system is reportedly designed to help differentiate between suspected Bangladeshi nationals and Indian citizens from linguistically similar regions, such as West Bengal or parts of Assam, where dialects often overlap. Officials have acknowledged that this is one of the project’s biggest challenges.
Chief Minister Fadnavis has publicly stated that the tool’s current accuracy stands at around 60 per cent, adding that it is expected to improve significantly with more data and testing over the next four to six months.
He has maintained that the AI will only “assist” officials and that final decisions will always involve document checks, human assessment, and legal scrutiny.
The project is estimated to cost around ₹3 crore and is being developed using state funding.
Government’s Rationale: Security and Governance
The Maharashtra government has justified the initiative by citing concerns over national security, misuse of forged documents, and pressure on public resources. State officials have claimed that illegal migrants often obtain fake identity papers to access welfare schemes, housing, and employment, making detection through conventional methods difficult.
Political leaders from the ruling dispensation have also linked the issue to broader law-and-order challenges, arguing that unchecked illegal immigration could strain social cohesion.
The AI tool, they say, is part of a wider strategy to modernise policing and governance through technology.
Officials have repeatedly emphasised that the system will not replace existing legal processes under the Foreigners Act and other relevant laws, and that deportation or detention decisions will continue to follow due process.
A Complex Context of Migration and Identity
The announcement comes amid heightened political focus on illegal immigration across several Indian states. In Maharashtra, the issue has gained renewed attention in recent months, particularly during election campaigning, with repeated claims about the presence of Bangladeshi and Rohingya migrants in urban centres.
However, experts warn that migration in eastern and north-eastern India has historically been fluid, shaped by economic necessity, displacement, and climate pressures.
India’s linguistic landscape is also deeply complex, with accents and speech patterns influenced by geography, class, education, and migration within the country.
Civil rights groups argue that relying on language as an indicator of nationality is fraught with risks in a multilingual society like India, where millions speak multiple languages or dialects that do not neatly correspond to political borders.
Concerns Over Bias, Profiling, and Misuse
The proposed AI tool has raised serious concerns among human rights advocates, legal experts, and technology ethicists. Critics warn that language-based identification could lead to profiling, false positives, and harassment of marginalised communities, particularly migrant workers and linguistic minorities.
There are also concerns about algorithmic bias – where AI systems trained on limited or skewed datasets may disproportionately flag certain groups. Once flagged, individuals may face repeated questioning, social stigma, or even detention, critics argue, even if they are Indian citizens.
Legal experts have pointed out that accent or speech analysis has limited evidentiary value in courts and could face judicial scrutiny if used improperly.
They have called for transparency about how the system works, independent audits, and clear accountability mechanisms.
The Logical Indian’s Perspective
Technology can be a powerful tool for better governance, but when it intersects with identity, migration, and citizenship, it demands the highest standards of care, transparency, and empathy.
In a diverse and multilingual country like India, language is not a fixed marker of nationality but a living expression shaped by history, movement, and survival.
While governments have a responsibility to address genuine security and administrative concerns, solutions must not deepen fear or reinforce suspicion against already vulnerable communities.
AI systems, especially those used in policing, must be subject to strict oversight, clear legal boundaries, and robust safeguards against misuse.
