Gurugram: A pavement that was never cleared of garbage showed up in official records as spotless, because the man responsible for clearing it simply asked artificial intelligence to erase the pile of garbage from the photograph.
The Gurugram Municipal Corporation Saturday terminated the services of four contractual employees, hired through the Haryana Kaushal Rozgar Nigam (HKRN), after uncovering what officials described as a mix of technological fraud and procedural violations.
Two of the sacked employees worked in the sanitation branch, and two in the property tax branch.
Municipal Commissioner Pradeep Dahiya told The Print that the property tax branch employees had failed to follow standard operating procedure (SOP) and were found in serious violation of departmental norms, while the sanitation branch cases involved a more brazen misuse of technology.
Also Read: In 1st budget under BJP mayor in 44 yrs, BMC hikes outlay to Rs 81k crore, focus on big-ticket infra
The AI-altered garbage photo
Among those dismissed was an employee named Wasim, tasked with resolving public complaints about uncleared garbage.
Instead of visiting the spot and clearing the waste, Dahiya said, Wasim ran the complaint photograph through an AI tool, digitally wiped the garbage heap out of the image, and sent the doctored picture to senior officials as proof to claim that the job was done.
Dahiya told The Print that the site where Wasim was found digitally wiping out the garbage heap was at Kherki Daula in Gurugram.
The second sanitation employee, Sonu, a resident of Jhajjar, took a different route to cut corners.
He allegedly exploited GPS spoofing technology to defeat the geo-fencing system meant to confirm an employee’s physical presence at the assigned location.
By spoofing his location, Sonu managed to register his attendance for duties in Gurugram while he was actually sitting at home in Jhajjar, Dahiya said.
Caught by system meant to be gamed
What undid both men, according to the Commissioner, was the corporation’s own monitoring software, which is designed to detect exactly this kind of manipulation.
“We have software that immediately tracks any misuse of AI and any changes made to metadata,” Dahiya told The Print, explaining that the alterations in Wasim’s photograph and the location mismatch in Sonu’s attendance logs were both picked up through metadata analysis.
The Commissioner added that the corporation does not rely on photographs alone to verify complaint resolution.
“Quick Response Teams (QRTs) are deployed to carry out random sample checks and physical verification at sites where complaints have supposedly been closed, a system that has increasingly caught discrepancies between what is reported and what exists on the ground,” he added.
Dahiya said the corporation has passed speaking orders in the matter, as mandated under government service rules, and forwarded its report to HKRN for further disciplinary action against the four employees.
He added that he had issued a stern message to all employees that any attempts to hoodwink the system would not be tolerated, and the person would lose their job.
(Edited by Amrtansh Arora)
Also Read: BMC in its grip, why BJP is now betting big on Bengaluru civic polls