Stopping the flow of fraud in government has always been a challenging task due to limited resources and human resources. But now, the onslaught of fraud has become an unprecedented fire station, made possible by AI-driven attacks. This escalation is so tense with already growing resources that it is important for government agencies to effectively prevent and mitigate fraud by adopting AI-powered tools. In fact, by tackling fraud, waste and abuse (FWA), governments can save 16% of their budget, according to a new global research report from SAS and Coleman Parkes.
As AI evolves, con artists will increasingly misuse it to expand sophisticated attacks and overwhelm traditional defenses. The interests are high: public trust, economic stability, global security. Governments must adopt AI-driven tools. Not only is it to maintain the pace, but to turn the tide. In this AI-powered arms race, governments must embrace advanced analytics and AI solutions and turn these technologies from fraudulent tools to resilience shields.
The rise of AI-driven fraud
According to the report, 95% of the global government agencies surveyed experience AI-powered FWA schemes. Scammers deploy AI-powered platforms to generate synthetic identities, create hyper-personalized phishing campaigns, and design malware that avoids detection. These tools analyze vast datasets to mimic human behavior, forge documents, and harness vulnerabilities in real time. Generated AI, for example, can create fake passports or invoices that are indistinguishable from real passports, allowing scammers to infiltrate the financial system and manipulate the procurement process.
AI also allows fraud to scale exponentially. Phishing campaigns can now deploy thousands of persuasive emails in seconds. Synthetic identity – constructed using actual manufactured data – bypasses traditional checks, allowing fraud to claim profits and open fraudulent accounts across jurisdictions. Meanwhile, AI-driven malware adapts to avoiding outdated security systems that surge before human-driven defense.
Deepfakes also appear in virtual meetings. As quoted in a recent Fortune article, AI was mimicking trustworthy colleagues.
The government's difficult battle
Faced with a surge in workloads, limited budgets and often outdated systems, how can government agencies face these threats? According to the report, only one in ten agents who have all the tools and resources needed to fight the FWA faces nearly a third of their large resource limitations. An outdated manual process allows scammers to create backlogs that are exploited.
Tax agencies say that when fraud exploits vulnerabilities, a large amount of real return filing, procurement fraud emits public funds, and citizens lose trust in the institution. Without AI-driven defenses, the government is reactive and always a step behind.
Fraud also hits the public's confidence. Of the 1,100 survey respondents, an astounding 96% of all public sector employees responsible for monitoring FWAs within the organization say the FWA has negatively affected citizens' trust in agents and their programs.
AI Defense Tool Kit
While it may seem like AI-powered con artists have the advantage, not being interrupted by things like regulations or law, they may not be long. The near future looks promising. The study shows that the use of machine learning for fraud detection is expected to expand from 36% to 84%. Furthermore, encouragingly, 30% of respondents currently use Genai solutions, while over 90% expect to use Genai over the next two years.
Advanced analytics and AI can play a pivotal role in this defense strategy. By combining data into a single dataset, software tools can analyze that large amount of data and detect anomalies and hidden patterns that indicate fraud. This enterprise approach allows for early and more accurate detection of fraud, minimizing false positives, reducing investigation costs and increasing inspector efficiency and productivity.
In payment fraud detection, the machine learning model combines behavioral profiling with rule-based detection in a layered fraud prevention approach. These models use advanced methodologies and statistical methods to identify dangerous transactions and flag them for further reviews while reducing legitimate user friction.
In law enforcement, analytics tools can connect different data and discover networks of fraud. For example, linking property records with benefits claims could reveal synthetic identity, but AI analysis of whistleblower hints can prioritize high impact research. This integration not only accelerates case resolution, but also ensures that fraud detection systems adapt to evolving threats.
AI ARMS Race Stakes
The market for generator AI is projected to reach $1.3 trillion by 2032. This growth is a double-edged sword. Technology has the unfortunate side effect of empowering fraudsters, but it can also equip governments with tools to dismantle schemes. Urgency is important. Institutions that slow the risk of AI adoption are increasingly overwhelmed by sophisticated attacks.
AI Arms Race has reached the field of fraud, and AI is now serving as the main tool for fraudsters to take advantage of vulnerabilities, with governments striving to keep pace. However, AI also offers solutions. It is an advanced platform that automates detection, discovers hidden connections, and acts in real time.
Governments now have to invest in AI-driven defenses or face higher costs of lost funds, erosioned trust, unmanageable risk. Luckily, this investigation shows that this is happening, with government agencies equipping themselves to fight back. In this race, the innovative and tactical applications of AI tools make the difference between leading the battle or falling behind.
Carl Hammersburg manages SAS's government and healthcare risk and fraud teams. Prior to that, he spent 20 years in anti-combust activities for Washington's Department of Labor and Industry, a comp insurance for monopoly workers. In 2004, CARL established its agency's comprehensive fraud program, covering tax and premium audits, claims investigations, provider fraud and collections.
Copyright©2025 Federal News Network. Unauthorized reproduction is prohibited. This website is not intended for users within the European Economic Area.
