Painting a Rosy Picture of AI, Despite its Dirty Reputation
/
The good, the bad and the ugly of artificial intelligence (AI) were presented during Monica Verma’s keynote session at the 2025 ACFE Fraud Conference Europe. In her presentation, titled “Humans, AI and Trust: Navigating the Modern-Day Landscape,” Verma, an award-winning AI, cybersecurity and leadership expert, started discussing significant years in the history of cybersecurity and AI. In the first quarter of 2023, when AI tools were released, a study showed that every two out of three employees at that time were using some form of generative AI tool. Unfortunately, there was also a 200 percent increase in voice-based scams that targeted the finance sector.
“We were seeing, in real time, a seismic shift happening,” said Verma. “We were seeing an evolution of cybercrime like never before.”
Verma said, in 2024, a trillion dollars were lost to digital fraud and scams alone. Meanwhile, synthetic identity fraud shot up 153 percent, cybercrime became a service and cyber-attackers could even be hired. Because the financial sector has the money, Verma said it has been an easy target for bad actors.
She explained that AI agents create intelligent outputs based on data sets that the technology has been trained on. In fact, AI only needs a 15-second audio clip to produce a voice clone, and it is becoming much harder for people to distinguish between what is real and fake nowadays. Because of this, the financial sector saw an 80 percent increase in fraud attempts, mostly driven by AI. With it already taking an average of 200 days before cyberattacks are discovered, AI is now making detection even harder.
Verma shared how people are also using AI for a lot of decision-making, even though they do not know how accurate and non-biased it is.
“An average adult makes around 35,000 decisions on a daily basis,” said Verma. “The question we really need to ask [is] how many of those decisions and what kind of decisions are being replaced or augmented by AI, and what does that mean for our society, for our businesses and individuals?”
Verma admitted that the first half of her presentation painted a dark picture of the widespread adoption of AI, but she said a lot of hope lies ahead for the technology, including how it relates to the work of fraud examiners. She said we need to think critically about AI and all the things it is good for and not good for.
Verma stressed that doing these five things will help fight fraud and build trust:
User awareness.
Recognizing that AI is great at analyzing patterns, data and behavior. Studies have shown that when machine learning is used on top of existing fraud prevention measures, fraudulent transactions are reduced by more than 40 percent.
Using AI to defraud and hunt scammers. You can use AI to detect deepfakes, use an AI chat box to beat scammers at their own game and more.
Utilizing AI for cybersecurity fundamentals. AI can help you detect if an email is suspicious, but you still need to practice cyber hygiene with secure passwords, authentication and more, and cyber and fraud teams must work together.
Practicing the responsible use of AI by keeping these five things in mind:
Are you using AI technology ethically?
Is it fair and non-biased?
Transparency: Why is AI making certain decisions, what factors was the decision made on and how did it come to that conclusion? This is known as “explainable AI.”
Inclusivity: Is the AI including all demographics?
Accountability: Who is accountable for AI’s actions? Is it the coder that created the tool, the company’s leaders, the users?
Following her presentation, Verma took questions from the audience. The very last question was probably in the back of the mind of many gainfully employed individuals as artificial intelligence gets smarter and more productive: When are robots going to take all our jobs?
Verma could not give a simple yes or no answer to this question, but she did say it is a positive that AI is helping lessen our loads so we can focus on other things. In her words, “Maybe we need robots to do the work that we are doing robotically today.”
