The Social Security Administration is working with federal partners to stop AI-powered fraud schemes, including deepfakes used to alter bank details and impersonate employees, before fraudulent payments are distributed. Losses from deepfake-driven fraud are projected to reach $40.1 billion by 2027, according to data cited by the agency's Office of Inspector General.

What Types of AI Fraud Is the SSA Actually Seeing?

Deepfake technology, which can realistically reproduce a person's voice and likeness, has become a tool for criminals targeting the SSA. The agency's investigators have documented several fraud schemes powered by artificial intelligence.

• Direct-Deposit Manipulation: Criminals use deepfakes to alter direct-deposit information, redirecting benefit payments to accounts they control.
• Impersonation Attacks: Fraudsters pose as other people or trusted organizations, sometimes invoking the names of real SSA employees to appear authentic.
• Social Engineering: AI-generated voice and video are used to convince SSA staff and beneficiaries to take actions that compromise security or enable fraud.

Chad Bungard, chief strategy officer in the SSA's Office of Inspector General, explained the scope of the challenge at a recent government event. The agency is treating this as an urgent priority because the financial impact extends beyond individual victims to the integrity of federal benefit programs.

How Is the SSA Preparing Its Workforce to Combat AI Fraud?

The SSA recognizes that fighting AI-powered fraud requires more than just technology; it requires people who understand the threat. The agency is investing in training and governance to ensure staff and investigators can recognize and respond to these emerging threats.

• Supervisor Training: The SSA is training supervisors who work with AI systems in any capacity, ensuring leadership understands both the capabilities and risks of artificial intelligence.
• Staff and Investigator Guidance: The Office of Inspector General is developing governance policies and training materials for frontline staff and fraud investigators.
• AI Governance Development: As AI takes on a larger role in both fraud schemes and fraud-detection tools, the agency is establishing clear policies for how AI should be used and monitored.

"SSA is training supervisors who work with AI systems in any capacity," stated Chad Bungard, chief strategy officer in the agency's Office of Inspector General.

Chad Bungard, Chief Strategy Officer, SSA Office of Inspector General

What AI Tools Is the SSA Already Using?

The SSA is not waiting passively for fraud to occur. The agency already uses specialized artificial intelligence tools to detect fraud and generative AI for administrative tasks. However, the agency is still determining how to approach agentic AI, which refers to AI systems that can operate autonomously to accomplish goals. In August, the SSA issued a request for information on an enterprise AI strategy that would cover its adoption of both traditional AI and agentic AI systems.

How Are Federal Agencies Coordinating the Response?

The SSA is not fighting this battle alone. The agency is part of a coordinated federal effort to prevent fraud before payments are made, rather than pursuing cases after fraud has already occurred.

• National Anti-Fraud Committee: The SSA participates in this committee, which aims to prevent fraudulent payments before they are distributed and coordinates federal efforts to examine evolving fraud risks, vulnerabilities, and possible mitigation steps.
• National Crime Defense Center Collaboration: The SSA's Office of Inspector General shares intelligence with the Justice Department's recently launched National Crime Defense Center and has embedded investigators and analysts to support its operations.
• National Fraud Detection Center: The SSA is a member of this interagency task force, which targets large-scale fraud against federal programs.
• AI Subcommittee: Bungard co-chairs an AI subcommittee with the SSA's chief AI officer, where the group discusses use cases under development, risk assessments, and the evolving AI landscape.

This multi-agency approach reflects a broader recognition that AI-powered fraud is a national security and public safety issue that requires coordination across the federal government. By sharing intelligence and embedding personnel across agencies, the SSA and its partners are building a more resilient defense against deepfakes and other AI-driven threats.

The projected $40.1 billion in losses by 2027 underscores the urgency of these efforts. As deepfake technology becomes more accessible and convincing, the SSA's proactive stance on training, governance, and interagency coordination may serve as a model for other federal agencies facing similar threats.