In a controversial move that’s raising red flags across the tech and public sectors, the Department of Government Efficiency (DOGE) is racing to migrate the Social Security Administration’s (SSA) legacy systems away from COBOL—within just months.
A High-Stakes Modernization Attempt
SSA’s infrastructure, which supports over 65 million Americans, is deeply rooted in COBOL, a programming language dating back to the 1950s. Despite its age, COBOL still powers critical operations like issuing Social Security numbers and calculating benefits. A safe and complete migration would typically take years. However, DOGE’s directive, led by Elon Musk ally Steve Davis, aims to accomplish this within a drastically shortened timeline.
Risking Millions of Lives for Speed
Experts warn that a rushed overhaul of a system this complex could result in widespread failures. “It’s not just about underpayment or overpayment,” said one SSA technologist. “The real danger lies in people not receiving their benefits and the system not even realizing it.”
This concern is amplified by the fact that the SSA’s core logic—everything from payment processing to benefit eligibility—is intricately coded in COBOL. A rushed rewrite without rigorous testing could lead to catastrophic errors.
A History of Delayed Digital Transformation
This isn’t SSA’s first attempt to modernize. In 2017, the agency outlined a five-year plan to phase out COBOL. That effort was sidelined during the 2020 pandemic. With DOGE now at the helm, the modernization is back—but at warp speed.
AI-Powered Translation: A Shortcut or a Trap?
Sources suggest DOGE may rely on generative AI to help convert tens of millions of lines of code into modern languages like Java. While AI can accelerate code translation, it often lacks the context necessary to ensure flawless execution, especially for mission-critical systems.
This raises significant parallels with the challenges explored in Inside Claude: What Anthropic’s AI Reveals About Creativity—and Deception, where the capabilities and limitations of AI’s decision-making were closely examined.
Internal Turmoil and Misinformation
DOGE’s aggressive tactics come amid growing political pressure on SSA, amplified by false claims from Elon Musk about fraudulent payments to “150-year-olds.” These claims, based on misinterpreted or outdated data, have intensified scrutiny of SSA’s operations.
Meanwhile, internal sources report an influx of inexperienced engineers—dubbed “DOGE operatives”—working within SSA. Their focus has largely been on verifying identities and rooting out perceived fraud, including launching the controversial “Are You Alive Project.”
System Integrity On the Brink
SSA’s architecture, including its custom-built MADAM database system from the 1980s, is a fragile web of interdependencies. One former senior technologist likened it to a game of Jenga: “Start pulling pieces out without a complete understanding, and the whole thing could collapse.”
Uncertain Timeline, Uncertain Outcomes
Despite the urgency, there’s no clear start date for the code migration. Internal planning documents don’t mention the shift, instead highlighting cost-cutting measures and AI integration for administrative tasks. With no roadmap and limited testing protocols, experts fear the worst.
As the SSA braces for one of the most ambitious digital transformations in its history, the stakes couldn’t be higher: the financial well-being of millions hangs in the balance.
