In what police are describing as India’s first case of its kind, investigators in Bengaluru say a 25-year-old accused in a triple murder consulted an AI chatbot at nearly every stage of a six-month plan to kill his live-in partner’s parents and sister, and have now written to the AI’s parent company seeking his complete chat history as evidence.
A Family Killed in Seegehalli
The murders took place on June 22 at a rented apartment in Bengaluru’s Seegehalli area, within the K.R. Puram police station’s jurisdiction, where Somasundar, 52, his wife Muthulakshmi, 48, and their younger daughter Supriya, 19, were found stabbed to death. Days later, police arrested the couple’s elder daughter Shweta, 24, and her live-in partner Kenneth, 25, both engineers by profession, after tracking them to Puducherry, where they had gone into hiding.
Investigators say that while both accused acted in tandem, Kenneth was responsible for the bulk of the planning and execution, and it is his alleged reliance on an AI chatbot, reportedly Google Gemini, that has drawn the case unusual attention. Police say he consulted the chatbot repeatedly over roughly six months, seeking guidance at various stages of the alleged conspiracy.
How Investigators Say the Chatbot Was Used
According to police, Kenneth avoided asking the chatbot direct questions about committing murder or disposing of a body. Instead, investigators allege he framed his queries as hypothetical scenarios, of the kind, “if something like this happens, what should someone do,” a pattern police believe was a deliberate attempt to extract relevant information without triggering the safety mechanisms most AI systems have built in for such requests.
Police say the resulting conversations touched on planning the killings, destroying evidence, removing bloodstains, and evaluating whether an iron furnace Kenneth had built for a planned cloud kitchen business could be repurposed for body disposal. A senior officer told The Indian Express that investigators had “almost considered naming the AI chatbot as an accomplice,” given the extent to which Kenneth appeared to rely on it, though police have been clear that they are not attributing any legal liability to the AI tool itself. Bengaluru Police have formally written to Google seeking the underlying chat logs to corroborate Kenneth’s own account, given during interrogation, of how heavily he depended on the chatbot throughout the planning period.
Financial Strain and Family Conflict at the Core
The alleged motive traces back to money and mounting family tension. Police say Shweta had borrowed around ₹50 lakh from her mother to fund Kenneth’s cloud kitchen venture, a business that investigators believe failed to generate meaningful returns, leaving the loan largely unpaid even as pressure to repay it grew. Kenneth reportedly told police during interrogation that it was ultimately a “trivial” dispute that pushed him toward the killings, though the underlying pressure had clearly been building for months.
Investigators say Shweta had separately expressed frustration with what she considered her mother’s controlling behaviour, a friction that had reportedly strained the family for some time, while psychological assessments during the probe found Kenneth had become severely socially isolated, relying almost entirely on Shweta for both emotional and financial support after leaving the IT sector.
What the Case Signals for AI Oversight
The episode has reopened debate around the safeguards built into mainstream AI platforms, which are generally designed to refuse or limit assistance connected to violent or illegal acts. Cybersecurity experts note that the case, if the allegations hold up, appears to illustrate how such safeguards can potentially be circumvented through indirect or hypothetical framing rather than explicit requests, a gap that developers of AI systems have been working to close as these tools see wider adoption.
Renowned cybercrime expert and former IPS officer Prof Triveni Singh said the case underscores how emerging technologies can be misused when combined with clear criminal intent, even as the same tools deliver genuine value across productivity, research, healthcare and education. He said AI cannot replace human judgment or legal accountability, and noted that digital evidence, spanning chat records, cloud data, device forensics and online activity logs, is becoming an increasingly central pillar of modern criminal investigations, one police forces across India will likely need to build far greater technical capacity to handle.
Police say the investigation remains ongoing as they await a formal response from Google and continue forensic verification of the digital evidence, with the precise role each accused played still being established before final charges are framed.
