Elon Musk wants to rewrite all the COBOL code behind U.S. Social Security—in just a few months

Elon Musk proudly wearing a DOGE t-shirt

What could possibly go wrong? — The Department of Government Efficiency (DOGE), now under the command of the ever-controversial Elon Musk since Donald Trump’s return to office, just keeps shaking up American institutions.

At the helm is Steve Davis and his squad of AI-fueled young mercenaries, storming through federal agencies with a mission to “modernize” the state infrastructure through lightning-fast overhauls and tech-messianic promises.

One of their latest obsessions? Completely rewriting the computer systems of the Social Security Administration (SSA), the historic backbone of U.S. social protection.

The catch: those systems are still powered by tens of millions of lines of COBOL, a programming language that dates back to the 1950s.

That doesn’t seem to faze DOGE, which plans to replace it with something more modern—like Java—within just a few months.

The announcement sent shockwaves through the tech community. Beneath the shiny tech slogans and promises of modernization, the project raises serious concerns about its feasibility—and, more importantly, its consequences.

Man in front of exploding building – Everything is under control, move along now

Testing? That’s for the weak.

Migrating a system of this scale is a Herculean task (at least, for anyone sane). Even under ideal conditions, such an overhaul would require several years of meticulous development, testing, and phased deployment.

Attempting it in just a few months poses massive risks to the system’s integrity—and could jeopardize benefit payments for the 65+ million Americans currently relying on them.

In an interview with WIRED, an SSA staffer warned of the dangers of such a rushed rollout: “One of the biggest risks isn’t just overpaying or underpaying beneficiaries—it’s not paying them at all, and not even realizing it. Invisible errors and omissions are a real threat.”

The SSA has already been under political pressure from the Trump administration. In February, Elon Musk falsely claimed that the agency was riddled with fraud, even pointing to alleged cases of 150-year-old beneficiaries.

In reality, these so-called anomalies stem from technical quirks in COBOL, which lacks a standardized date format.

Some incorrect entries in the system show birthdates as May 20, 1875—a placeholder date used by convention in certain COBOL systems.

COBOL = GOAT?

COBOL remains a cornerstone of global financial systems. Its reliability and ability to handle massive data volumes make it a vital—if invisible—component of countless public services and critical infrastructure.

The COVID-19 pandemic in 2020 had already highlighted the massive dependency of U.S. institutions on this ancient language—with very real consequences.

The state of New Jersey, overwhelmed by unemployment claims, ran into an unlikely bottleneck: a COBOL developer shortage.

So much so that the governor publicly called for volunteer COBOL programmers to urgently get the systems back online.

Similarly, about 95% of ATM transactions worldwide are believed to run on COBOL. The SSA alone handles nearly 840 million financial transactions annually, totaling over $1.3 trillion in benefits paid in 2023.

Experts are sounding the alarm: even minor migration errors could cause catastrophic failures.

Take decimal handling, for instance—it’s very different in COBOL compared to more modern languages like Java. A rounding mismatch could trigger widespread payment errors.

AI-generated image of Elon Musk at a desk with sunglasses, behind a DOGE sign

All that… for what, exactly? Because despite its age, the current system still fulfills its primary mission: delivering social benefits to tens of millions of people.

Given that, DOGE’s enthusiasm feels less like a rational modernization effort and more like a PR stunt chasing yet another flashy headline.

In fact, the SSA already attempted a similar modernization in 2017 with a five-year plan. That initiative ultimately failed—partly due to the pandemic, and partly due to the lack of a clear plan.

So far, no detailed migration strategy has been presented. All we know is DOGE’s stated intent—true to form—to rely on generative AI to speed things up.

Also on The Coding Love: