WASHINGTON, D.C. – Today, U.S. Sens. Sherrod Brown (D-OH) and Bob Casey (D-PA) pressured the Trump Administration to provide answers to the American public following a report from the New York Times this week that the White House is working with the Social Security Administration (SSA) on a proposal to monitor social media accounts of Americans who claim disability benefits. The Senators’ letter raises several questions about the proposal related to the privacy of American citizens, the already limited resources available to SSA workers, the use of artificial intelligence, and plans to increase the scope of the social media monitoring program.
“Rooting out fraud in all federal programs is an important goal, which we have long supported. However, we are concerned that this proposal – which involves the federal government spying upon taxpayers – is riddled with problems related to due process, privacy, the prudent use of valuable, yet limited Social Security Administration (SSA) funds, and the assumption that people with eligible disabilities can be determined to be committing fraud through their social media posts. The unspoken assumption of this social media analysis is that people with disabilities are inactive, pitiful shells who have no social life and no interaction with others. It is an inherently disrespectful and demeaning set of assumptions that marginalizes and minimizes their contributions to society and their lives,” the Senators’ wrote.
The letter outlines several potential problems with a social media monitoring program, including the inaccuracy of social media posts, the costs of training staff to monitor posts for an agency that has already experienced deep budget cuts, and the potential for abuse of the program.
The Senators’ believe the Administration should answer questions and provide transparency to taxpayers before undertaking such a serious, and potentially harmful, proposal.
A copy of the letter can be found here and below.
Dear Director Mulvaney and Acting Commissioner Berryhill:
We are deeply troubled by the recent New York Times Article: “On Disability and on Facebook? Uncle Sam Wants to Watch What You Post.”
This article addresses how the “Trump administration has been quietly working on a proposal to use social media like Facebook and Twitter to help identify people who claim Social Security disability benefits without actually being disabled.”
Rooting out fraud in all federal programs is an important goal which we have long supported. However, we are concerned that this proposal – which involves the federal government spying upon citizens – is riddled with problems related to due process; privacy; the prudent use of valuable, yet limited Social Security Administration (SSA) funds; and the assumption that people with eligible disabilities can be determined to be committing fraud through their social media posts.
Eligibility for SSDI benefits is determined through a rigorous medical review process. The process can take up to two years and includes review of extensive medical evidence. The assumption that the process can be negated by a social media post is patently ludicrous. The assumption also ignores that people with disabilities can and do have full lives that include social events, vacations, travel, and other activities. People with disabilities are vibrant, engaged, social individuals and should never be judged on the activities in which they participate. The unspoken assumption of this social media analysis is that people with disabilities are inactive, pitiful shells who have no social life and no interaction with others. It is an inherently disrespectful and demeaning set of assumptions that marginalizes and minimizes their contributions to society and their lives.
In an effort to better understand this proposal, we appreciate your attention and responsiveness to a number of questions regarding 1) fairness to citizens, 2) impact upon SSA workers, 3) use of artificial intelligence (AI), and 4) the scope of the Administration’s social media surveillance plans.
Fairness to Citizens: Throughout the disability application process, the burden remains on the claimant to establish that they are unable to work. Social media posts are discrete snapshots that show an individual at a particular, specific moment. The fact that a person spent a few minutes dancing, or fishing, or holding a grandchild has little bearing on what work activities they could carry out on a sustained basis, which is what disability adjudicators are trying to determine. ALJ review of social media posts does not provide them with substantial evidence showing that a claimant is able to work. To that end please answer the following questions:
Impact Upon SSA workers: SSA employees are increasingly stretched thin – being asked to do more with less. In fact, since 2010 the SSA has lost 12 percent of its staff and faced a 9 percent decline in its budget. During the same time period, SSDI beneficiaries increased to over 11 million and the number of retirement and survivors insurance beneficiaries has increased. The result is entirely predictable – SSA offices have closed, wait time for calls have increased, and the number of workers deciding whether applicant disabilities warrant disability insurance (DI) has decreased.
Disability examiners are already stretched thin. Directing SSA employees to follow the social media habits of over 10 million SSDI or SSI beneficiaries and then investigate what they find is a costly undertaking shifting resources away from other key priorities. Furthermore, directing Administrative Law Judges (ALJs) to perform such an analysis is an even bigger waste of time. The SSA spends about an eighth of its administrative funds on program integrity activities and has a large staff at the Office of the Inspector General who work to ensure program integrity. Meanwhile, there are over 800,000 people waiting for a disability decision, including tens of thousands who had their hearings and are waiting for decisions. The ALJs should focus on that work, which only they can do, and let the investigators investigate credible evidence of fraud. Spending time looking for people on Facebook or Instagram is not a good use of ALJ time.
To that end please answer the following questions:
Use of artificial intelligence: If SSA is not proposing to increase its staffing budget or planning to alter its administrative priorities, there is no practical way for SSA employees to efficiently monitor social media activity. This leads to a troubling option: the use of artificial intelligence (AI) to monitor citizens. Recent news reports have detailed how authoritarian regimes are using AI to mine personal data and track their citizenry. Such an approach is antithetical to our nation’s understanding of privacy, separation of powers, and individual liberty. In fact, the benefit of whatever sums AI applications help recover from SSDI fraudsters is outweighed by the cost to system of ordered liberty. To that end please answer the following questions:
Scope of the Administration’s Social Media Surveillance Plans: We have deep and broad concerns about the use of social media surveillance to ferret out federal program fraud. Moreover, we are deeply concerned that the administration might expand this surveillance-state tactic so that there is social media monitoring of: veterans claiming VA disability benefits, farmers using USDA conservation programs, or entrepreneurs operating a pass-through corporation. To that end, please answer the following questions:
We appreciate your prompt attention to the matter and your detailed responses. Please provide our offices with a detailed, written response by March 29, 2019.