As of September, the Trump administration is considering a new proposal that if implemented, could effectively identify mentally ill Americans as potential enemies of the state, according to disability advocates. Per a September 9 report in the Washington Post, the White House was briefed on a proposal for the Stopping Aberrant Fatal Events by Helping Overcome Mental Extremes project a.k.a. SAFEHOME, a project that would collect biometric data from smart devices including Fitbits, Apple Watches, Google Homes, and Amazon Echos to look for “neurobehavioral signs” of “someone headed toward a violent explosive act.” The proposal came shortly after the August 2019 mass shootings in El Paso, Texas, and Dayton, Ohio, and unsurprisingly, offered a way to redirect attention from the president’s reluctance to enact common-sense gun reforms such as universal background checks and closing the gun-show loophole.
The proposed program, assembled at the request of Ivanka Trump and championed by former NBC chairman and Trump pal Bob Wright, is based on the idea that technology might predict mass shootings before they happen. More specifically, it’s grounded in the belief that monitoring mentally ill people is the most viable route to preventing such events, a belief that President Trump espoused in the wake of the August shooting when he publicly stated that “mental illness and hatred pull the trigger, not the gun.” Experts in mental health and violence prevention, however, say that both the president’s statements and the SAFEHOME proposal are based on the false premise that mental illness and mass shootings are directly linked—a claim that has been repeatedly debunked by research.
SAFEHOME is part of an encompassing initiative to establish a new agency called the Health Advanced Research Projects Agency (HARPA). The new agency would be modeled after the Defense Advanced Research Projects Agency (DARPA), the Department of Defense’s multi-year project to develop weaponry and other technology for the U.S. military. Much of what DARPA has worked on in its 62 years as an agency is cloaked in national-security secrecy, but projects we do know about range from souped-up versions of existing technology (ballistic missile defense, the M16 assault rifle) to sci-fi–novel material (robot insect spies, a brain-computer interface program that uses brain scans to identify potential threats to soldiers in the field). In other words, the Trump administration wants to take an agency model specifically designed to identify and thwart terrorists and point it directly into American homes—a plot point that seems ripped from the pages of George Orwell’s 1984.
On Twitter, Sherrilyn Ifill, president of the NAACP Legal Defense and Educational Fund, pointed out that this is an idea that should terrify all Americans—and is also incredibly absurd. “The proposed data collection goes beyond absurdity when they mention the desire to collect FitBit data,” George David Annas, deputy director of the Forensic Psychiatry Fellowship Program at SUNY Upstate Medical University, recently told Gizmodo. “I am unaware of any study linking walking too much and committing mass murder. As for the other technologies, what are these people expecting? ‘Alexa, tell me the best way to kill a lot of people really quickly’? Really?”
There’s also the issue of consent, which has increasingly become a key focus in conversations about technology and privacy, and remains crucial as companies continue to exploit a lack of social-media literacy. Though proponents of SAFEHOME point to the project’s focus on volunteer data, arguing that Americans would be given the choice of opting in to the project, the 2016 election and Cambridge Analytica’s use of millions of Americans’ Facebook data showed us that “volunteer data” is an imprecise term. When user data is controlled by companies that have collected it—and that aren’t required to disclose how it will be used—the question of whose data is willingly “volunteered” looms large.
Beyond the project’s lack of basis in both reliable science (and, well, reality), one of the program’s biggest risks is that it’s guaranteed to generate large numbers of false positives. Regardless of the risk factors assessed, SAFEHOME would identify far more innocent, nonviolent people as risks to their communities than it would potential mass shooters. And how the federal government chooses to use or share that information would have real-world implications for millions of Americans. Of course, risk-assessment measures wouldn’t be applied evenly to all Americans. Imagine a world in which you are denied a loan or prevented from boarding an airplane with your family because Alexa decided that you yell too much. Those are the sorts of implications a program like SAFEHOME could carry.
This isn’t just speculative: Hundreds of schools, banks and hospitals have already installed “aggression detectors”—microphones equipped with algorithms designed to detect anger and stress prior to violence erupting—despite recent reports of the devices that are falsely triggered by children cheering for pizza or playing Pictionary. The U.S. government has always been more comfortable with anger from some groups than others, and the stakes of using aggression detectors and similar measures could even include the mass institutionalization of Americans determined by SAFEHOME to be high risk, if the president’s recent statements are to be believed.
Indeed, Trump has already suggested, repeatedly, that a key component of his administration’s plan to combat gun violence is rebuilding mental asylums to “deal with” these “sickos.” “These people are mentally ill and nobody talks about that. But these are mentally ill people, and people have to start thinking about it,” Trump told reporters in August. “I think we have to start building institutions again because, you know, if you look at the ’60s and ’70s, so many of these institutions were closed… But a lot of our conversation has to do with the fact that we have to open up institutions. We can’t let these people be on the streets.”
The president’s comment referred to the closure of large-scale mental institutions in the 1960s that followed the signing of the Community Mental Health Act by President John F. Kennedy. The intent of the law was to move the nation toward a community-based model for mental healthcare and research and away from mass institutionalization. Trump has acknowledged the financial factors that contributed to the decision to close such institutions, but neglects to mention that the law was also enacted to combat widespread abuse and neglect of patients who often spent their lives warehoused in mental institutions. Such abuse remains endemic: As BuzzFeed News reported in 2016, more than 4,000 cases of abuse and neglect were documented in a single year in the president’s own home state of New York, including 37 incidents that resulted in a patient’s death.
It’s not that a program like SAFEHOME shouldn’t exist: In an ideal world, it could focus on identifying Americans in need of help and connecting them with mental-health resources. But even in studies of suicidal behavior—an area with more concrete data and knowledge than behaviors like gun violence—it is extremely difficult to predict which participants are in need of help. In one recent study covered in Slate, researchers were able to accurately predict only 1 percent of cases in which a patient was suicidal. Identifying people who are likely to commit mass shootings is an even harder task, and we just aren’t there yet in terms of technology and understanding. Instead, we risk a system that only further marginalizes marginalized groups by providing yet another excuse to keep a “close eye” on people of color and people with mental illnesses.
We risk a system that only further marginalizes marginalized groups by providing yet another excuse to keep a “close eye” on people of color and people with mental illnesses.
By far, the most reliable risk factor in predicting mass shootings is gender: The overwhelming majority of mass shootings are carried out by men; several of the recent mass shootings in the United States were revealed to be racially motivated and carried out by white supremacists who explicitly referenced the inflammatory rhetoric of Trump’s own supporters. It’s hard to envision a proposal that advocates monitoring all white men, despite the fact that that such profiling would have far more basis in statistical reality than the proposed scapegoating of mentally ill Americans.
Ultimately, that’s all SAFEHOME is—scapegoating. It’s a four-year, 40-to-60–million-dollar attempt to distract Americans from the fact that Donald Trump is in the pocket of the NRA, all at the expense of mentally ill Americans and the American taxpayer. And, as we continue to allow the epidemic of gun violence to rip through our nation, it’s certain to come at the expense of more innocent lives.