by Countable | 11.17.17
The Department of Homeland Security plans to screen immigrants and visa applicants with predictive "extreme vetting" software.
"Over the summer," the AP explained, “DHS published a ‘statement of objectives’ for a system that would use computer algorithms to scan social media and other material in order to automatically flag undesirable entrants — and to continuously scan the activities of those allowed into the U.S.”
The project, dubbed the "Extreme Vetting Initiative," would use artificial intelligence (AI) to help determine whether an immigrant “intends to commit criminal or terrorist acts” or is more likely to become a “positively contributing member of society.”
On Thursday, an alliance of 54 computer scientists, engineers, mathematicians and other experts in automated decision-making sent a letter to DHS expressing "grave concerns regarding Immigration & Customs Enforcement’s (ICE) proposed ‘Extreme Vetting Initiative.’"
Also on Thursday, a coalition of civil rights groups launched a petition urging IBM to avoid working on the program.
In their statement of objectives, DHS wrote that the purpose of the Extreme Vetting Initiative is to "establish an overarching vetting contract that automates, centralizes and streamlines the current manual vetting process while simultaneously making determinations via automation if the data retrieved is actionable."
"ICE," they wrote, “must develop processes that determine and evaluate an applicant’s probability of becoming a positively contributing member of society as well as their ability to contribute to national interests in order to meet the [Executive Orders] outlined by the President … including ‘Protecting the Nation from Foreign Terrorist Entry into the United States’.”
In their letter to DHS, the dissenting researchers called the plan "neither appropriate or feasible."
"Simply put, no computational methods can provide reliable or objective assessments of the traits that ICE seeks to measure. In all likelihood, the proposed system would be inaccurate and biased."
The technology experts – hailing from academia and tech firms such as Google and Microsoft – said that current AI isn’t capable of determining whether an individual will "contribute to the national interests" or become a “positively contributing member of society.”
"Neither the federal government nor anyone else has defined, much less attempted to quantify, these characteristics," the researchers wrote. “Algorithms designed to predict these undefined qualities could be used to arbitrarily flag groups of immigrants under a veneer of objectivity.”
However, Jacob Shapiro, a Princeton University politics professor and president of investigative analytics firm Giant Oak, told AP that the effectiveness of the vetting system would depend on how it’s crafted.
"There are many ways one could meet that statement of objectives, some of which would make the system fairer, more equitable and faster for those seeking to immigrate," Shapiro said. “But some of which could be biased and unfair, as any algorithm can be.”
Shapiro had joined representatives from IBM, LexisNexis, Deloitte and others at DHS’ summer informational meeting about the Extreme Vetting Initiative. It was IBM’s attendance at that meeting that led to recent calls for a boycott against the tech company.
IBM said this week that its attendance was routine, and not indicative of plans to work with DHS in developing a program.
"As there is no active project being proposed, it is premature to say whether there is even an opportunity that IBM would consider pursuing," said IBM spokesman Ian Colley.
Christopher Padilla, IBM’s vice president of government affairs, went further, saying in an email that the company "would not work on any project that runs counter to our company’s values, including our long-standing opposition to discrimination against anyone on the basis of race, gender, sexual orientation, or religion."
Regardless of the eventual outcome of the petitions and letters, President Donald Trump remains committed to his campaign promise of "extreme vetting." Following the October 31 truck attack in New York City, Trump tweeted:
I have just ordered Homeland Security to step up our already Extreme Vetting Program. Being politically correct is fine, but not for this!— Donald J. Trump (@realDonaldTrump) November 1, 2017
Are there "many ways one could meet that statement of objectives"? Or do you believe “no computational methods can provide reliable or objective assessments of the traits that ICE seeks to measure”? Hit Take Action, tell your reps, then share your thoughts below.
— Josh Herman
(Photo Credit: blossomstar / iStockphoto)
Written by Countable
Using AI to arbitrarily determine if someone should be allowed to enter our country is another example of this administration pushing their responsibilities elsewhere (the generals are now the commanders, Congress is responsible for the ACA not being repealed and Obama is responsible for everything else). I spent 40 years in the tech industry and know that computers are not the answer to everything. Sometimes you need a human to make a decision. Have we already forgotten the algorithms that made it easy for Russian bots to buy ads on Facebook during the last presidential election? If there had been human intervention, maybe the disinformation campaign would have been less destructive. This is NOT extreme vetting -- it is another case of dereliction of duty.
As someone who has worked in tech for over a decade, is married to an immigrant, and works with kids in the system as a GAL, the idea of "extreme vetting" is a joke. Going through immigration in any way, from any country involves an amazing amount of background scrutiny on every level. It takes months if you are lucky enough to have the money for a good lawyer, years if you are not. I always notice that those who are most concerned with how "easy" it is to legally immigrate to the US and keeping out the undesirables are also people who don't know very basic facts about how our big, bulky, expensive, mistake prone immigration system works. For those who are afraid of people who aren't just like them, immigration is a great place to create a false sense of security for themselves. These are the same people who see mass shooters that DO resemble them far more often than not (white, christian, conservative, legally gun-owning men), yet they are somehow still more worried about Muslims - ALL Muslims - than they are concerned with the fire in their own back yard. AI and machine learning is in its infancy. How many US citizens would be considered productive enough, or useful enough, or patriotic enough if *their* Facebook and Twitter feeds were parsed?
If the VP of IBM has concerns over the ethics of a program, then that should be cause for some definite alarm bells here at home.
How hard would it be for this vetting program to delve into U.S. citizens daily affairs. We should believe that this would be prohibited by law, how often has an intelligent agency or arm of the government abused the systems that can monitor confidential activities? How well would a system like this work? Face Book, Google and how many others have delved into this method of voting data and we know that they have failed. These companies continually complain that just a minor adjustment in programming algorithms will fix it but these adjustments never cease. This proposal though is dealing with people’s lives, yes they may not be U.S. citizens but I’m use their lives are just as important to them as ours is to us and they deserve to deal with a human and a human needs to make any decisions made for responsibility and authority to be assigned. Computers may assist but the final decision must be made by a human and that decision should override any computer’s input.
This is worse than Big Brother in 1984. Wow, so now a computer will be profiling us. Developing killer robots to disperse unruly Resistance crowds next? What the actual fuck is going on in your heads? How can you even think this is a good idea?
All the terrorists were Muslims, so eliminate them as immigrants. They do not want to assimilate into our culture. They come here with the intent to transform America into a Muslim culture. We don’t want their kind.
This program is ridiculous!!!! This system will be extremely biased and we all deserve better.
If the state of the art in AI can guide driverless vehicles safely through city traffic, it stands to reason that AI can also aid in centralizing and streamlining the data intensive current manual process of vetting immigration applicants. I’m further concerned that protests against extreme vetting aren’t as much concerned with efficiency and accuracy as they are with disabling entirely all vetting against terrorist threats.
It is time that we take the strongest vetting program we can get. I am tired of seeing these long wolfs kill Americans. If we don’t do more we will be like Europe. Please vote yes
This sounds like a great place to start, and may provide information about who a person really is before considering allowing them into our country. i cannot imagine the thought process of those opposing this... all the PC retoric needs to end and people need to wake up to reality.
Unacceptable to anyone who loves America. That this Administration is unable to keep citizens safe without ICE, walls, extreme vetting is alarming, but the possible uses of this software is frightening.
ICE extreme vetting is wrong. Experts in the field writing as much to Department of Homeland Security. Must be stopped
There are enough red flags going up by in the know people that is cause for great concern. I do not think this system is set up to fairly assess anyone wanting to immigrate here. Do not be hasty to keep campaign promises and make us regret what is put into practice. I oppose this hasty inaccurate database.
ICE should be put on ice and disbanded! They are bounty hunters disguised as Federal Agents, and they are the ones who should be under extreme vetting! They are breaking up families and separating underage children from their parent(s)!
This idea is appalling. Consider just the single implication of names. No one is the only person with their name. AI is extremely unlikely to catch that mistake every single time. Lives are at stake. There are myriad other reasons why this is a dreadful idea; computer glitches, hacking, user error. This is much too important to trust to “an app”!
This is not actionable and extremely dangerous. AI is in its infancy and there is no way to determine who will be a contributing member of society and what does that even mean? Social contributions take many forms and the govt is going to define it? God help us. Trump isn’t a contributing member of society. Would he be flagged? And the thing with terrorism. What are they expecting? We don’t make those requirements of anyone. This sounds like an end run around racism. Skew it against Muslims and persons of color. This is an ignorant, ill conceived proposition from an admin that doesn’t believe in science. Apparently, they do believe in self-created science or FAKE science. Now we’ll hear that he knows more about science than Noble prize winners.
This is s miserable, white supremacist, Nazi program is the worst immigrant program since the McCarthy era and I couldn’t say it better than: "Simply put, no computational methods can provide reliable or objective assessments of the traits that ICE seeks to measure. The proposed system would be inaccurate and biased."
I say, I may WARM to the idea....ONLY IF, Trump and his swamp monsters are the first to use it, if AI deems them to not be productive members of society, then MAYBE. BUT if Trump and his cabinet are allowed to comeback, NOOO WAY. Not sure why that big dummy thinks that AI, can factor the ignorance of his base. Because as they get all hot and bothered and alienating people like they are of this continent to begin with. Because of the "extremist attacks" lately they seem to be of the right wing Christian extremist. If this is what PATRIOTS claim, "is thanks to American progress" that wiped out my ancestors, the First Nations people... I can see how and why they feel like they're Under attack. It is mainly due to being Insecure, and dumb. That way they can blame a stupid computer... Smh
This software would deny my entry to the United States based on my views expressed on countable. I am 36 years old white and male I am employed in a trade and pay taxes.
Software just lets you automate your preconceived ideas and makes it easier to discriminate.