“Moral and accountable AI for public good is our core motto,” mentioned Maharashtra Chief Minister Devendra Fadnavis. “AI has the facility to rework by enhancing effectivity, enhancing high quality of life and delivering true ease of residing for each citizen.”
The state even created a particular goal car to introduce AI options in legislation enforcement. It’s recognized by the acronym MARVEL, which stands for Maharashtra Superior Analysis and Vigilance for Enforcement of Reformed Legal guidelines.” The imaginative and prescient behind creating MARVEL is to construct a platform that companions with international deep-technology leaders to develop AI copilots that may essentially remodel the way in which we govern,” mentioned Fadnavis.
Amid a surge lately of investing scams, refund frauds and digital arrests, Nitu’s case is simply one of many tens of millions of cybercrime complaints lodged in India. In accordance with I4C knowledge cited in a parliamentary reply on Dec. 2, 2025, practically 2.27 million cybercrime circumstances have been reported in 2024 alone. Greater than 228 billion rupees (US $2.5 billion) have been misplaced to cyber criminals that 12 months. I4C is the nationwide company arrange by the Ministry of Dwelling Affairs to create a framework for stopping, detecting, investigating and prosecuting cybercrime in India.
With most investigators not well-versed in tech, having an AI-powered digital assistant readily available has eased their workload.
In Nitu’s case, lodging her felony criticism – often called a primary info report or FIR – took simply quarter-hour on the Saoner police station in Nagpur.
She provided the rip-off website’s social media account and URL, cell numbers, screenshots of her textual content messages and financial institution statements. All this info, be it on a PDF file or a handwritten be aware in English, Hindi or native language Marathi, could be uploaded, extracted and filed of their respective sections by MahaCrimeOS AI in minutes, liberating an investigator from tedious handbook work to create a case file.
Assistant Police Inspector Ashish Singh Thakur, who dealt with Nitu’s case, mentioned that earlier than it might have taken two to 3 months to assemble info for the investigation.
“We now have to go to a number of banks as a result of the funds have been transferred throughout greater than 40 to 50 completely different banks … We now have to jot down them a letter and take an announcement from them. Then, there have been a number of cell numbers used … I’ve to get the IP addresses. Numerous handbook work,” he defined.
As an alternative, it took Thakur a couple of week to finish these duties and listen to again from these service suppliers.
“With a number of clicks I can request the decision element information of a number of numbers. I don’t should draft the letters; they’ve been drafted by MahaCrimeOS AI. And I don’t have to fret about lacking out info or making errors in my letters, and having to start out over again,” he mentioned.
He can now deal with seven to eight circumstances a month, in contrast with only one beforehand.
Instruments to assist legislation enforcement
Automating these frontline processes was only one a part of the problem, mentioned Ram Ganesh, founder and CEO of CyberEye, the impartial software program vendor (ISV) that constructed the engine powering MahaCrimeOS AI. The corporate labored with MARVEL to customise it for Maharashtra police.
CyberEye is a Microsoft Accomplice ISV and works carefully with the Microsoft India Improvement Heart.
“Our first problem was there was very minimal standardization,” he mentioned.
This ranged from utilizing various phrases in reviews to explain the identical monetary fraud or felony methodology, to not having commonplace authorized notices searching for info from banks.
This made it tougher for the police to determine on subsequent steps, uncover if circumstances have been linked or construct up legacy data.
CyberEye labored with MARVEL to include the Maharashtra police’s investigation protocols and make the platform out there in Marathi, for ease of use by investigators.
They then added AI, within the type of an investigation copilot that analyses the case info, generates an automatic workflow and suggests investigation pathways. The copilot was constructed utilizing Azure OpenAI Service in Microsoft Foundry and secured with Defender for Cloud.
The platform additionally has built-in entry to India’s felony legal guidelines and open-source intelligence, which can be utilized to hyperlink crimes and find suspects.
A key benefit of the platform is its capacity to adapt present best-practice protocols, add a stage of information particular to a criminal offense after which develop tailor-made investigation plans, mentioned Harssh A Poddar, superintendent of police, Nagpur Rural, and CEO of MARVEL.
“Now you’ve a copilot that’s ready to try this work taking over board data sources that exist throughout the Maharashtra police,” he mentioned.
The system additionally frees officers to focus extra on crime-solving and sufferer help, by replicating and automating administrative duties they do in a typical day.
When an investigator logs on to MahaCrimeOS AI, they’ll view all of the circumstances they’re engaged on, whether or not there may be new info on it or motion to be taken.
In a monetary fraud case, such actions embrace asking a financial institution to freeze an account, requesting a telephone quantity be blocked or {that a} social media account be taken down.
In the meantime, if name element information have arrived from a telecommunications firm, they’ll ask Investigation copilot easy methods to proceed. Or they’ll click on on its suggestion to research the information or collect open-source intelligence on a telephone quantity.
These options allow officers to develop a really individualized investigation plan for each single cybercrime, mentioned Poddar.
