Content provided by Cognyte
By Pamela Benke, public safety data & AI strategist, Cognyte
Criminals are weaponizing generative AI to unleash realistic-looking phishing campaigns and deepfake impersonations, roiling police bureaus nationwide. The combination of fake, AI-generated video and web content with voice cloning technology has made it easier than ever for criminals to assume a disguise, commit fraud and cover their tracks.
Today, these sophisticated capabilities are accessible to anyone with a smartphone.
Police investigators are feeling the strain. The Identity Theft Resource Center reported a 148% rise in impersonation scam cases between April 2024 and March 2025. In most of these cases, criminals impersonated a business or financial institution representative, easily fabricating legitimate-looking communications and web content to deceive victims.
Flood of fake data
Deepfake content and scams are proliferating rapidly, and often below the radar because many of these scams aren’t reported by victims for fear of additional reputational damage. The number of cases that have been reported are staggering. According to data from the World Economic Forum, cases of deepfake-enabled fraud in North America increased by approximately 1,740% between 2022 and 2023.
Local and state police departments are hard pressed to make a dent in these fraud cases, partly due to the sheer volume of fake data being generated for criminal enterprises. Amid the deluge, crime solving rates are dropping sharply and case resolutions are increasingly harder to achieve.
Entire fake websites and social media profiles can be fabricated with pushbutton ease. The torrent of multimedia fake data is largely unstructured and therefore extremely difficult – if not impossible – to efficiently organize and analyze with manual techniques.
When AI-powered scams reach the media headlines, the reputational damage done to victims is amplified yet again. This puts more pressure on police to quickly identify and resolve fraud before it becomes a major story on the local news.
The common citizen isn’t equipped to defend themselves from deepfake fraud. They don’t know what to look for. Buffeted with fake communications from fake sources, they don’t know who to believe.
They count on the police to maintain law and order and public safety. Absent that trust, anxiety and unrest will only grow. This can manifest in increased crime and caseloads. Meanwhile, the deluge of fake data only continues to grow, obscuring serious crimes in the murk.
When case resolutions and justice itself grow elusive under the strain of illicit AI content, police reputations will suffer. There’s an urgency to act now.
Investigative GenAI levels the playing field
It’s time to fight fire with fire.
Police forces can solve cases faster and uncover the misuse of AI with an equally powerful application of AI for good. Dedicated GenAI tools – designed solely for the purpose of law enforcement and investigation – have arrived to ease the burden.
It’s exponentially easier to identify and neutralize AI deepfake threats and scams when the sifting and analysis of massive amounts of investigative data can be automated reliably and efficiently.
Crucially, new investigative GenAI tools are easy and intuitive to use, eliminating complexity and speeding case resolutions. The ability to explore and harness huge volumes of investigative data with natural language prompts brings the ability to get quick, accurate answers to complex questions spanning multiple data formats.
Imagine executing the digital forensics on a confiscated smartphone, for example – with thousands of stored photos, videos, recordings and documents – and then crunching that data into a tight summary in seconds. Analysts can review reports at a glance and prioritize and focus their investigations quickly and with clarity, without wasting time on false leads.
Taken to another level, a police investigator could inquire if this confiscated smartphone is connected to other open cases. An investigative GenAI assistant can very quickly compare available data sources, such as communication records, open-source data and automatic license plate recognition (ALPR) data, to show hidden links between the confiscated smartphone, a stolen vehicle and fraudulent social media accounts.
With these capabilities, unusual patterns can be surfaced within minutes, reducing manual work and helping investigators act faster.
Investigative GenAI tools can help investigators independently verify, explain and contextualize findings using established police methods, with end-to-end transparency and integrity.
GenAI – applied for good – can reduce repetitive, time-consuming tasks, bring structure to unstructured data, surface hidden insights and ensure investigators spend more time exercising judgment, interviewing witnesses and making decisions that require human intuition.
It levels the playing field and puts police forces a step ahead of adversaries using sophisticated tools. It can enable faster case resolutions, faster time to justice and faster relief for fraud and scam victims.
The time to act is now, before the deluge of fake data becomes impossible to navigate.
For more information, visit Cognyte.