The Crime You Have Not Yet Committed
Computers are getting pretty good at predicting the future. In many cases they do it better than people. That’s why Amazon uses them to figure out what you’re likely to buy, how Netflix knows what you might want to watch, the way meteorologists come up with accurate 10-day forecasts.
Now a team of scientists has demonstrated that a computer can outperform human judges in predicting who will commit a violent crime. In a paper published last month, they described how they built a system that started with people already arrested for domestic violence, then figured out which of them would be most likely to commit the same crime again.
The technology could potentially spare victims from being injured, or even killed. It could also keep the least dangerous offenders from going to jail unnecessarily. And yet, there’s something unnerving about using machines to decide what should happen to people. If targeted advertising misfires, nobody’s liberty is at stake.
Critics are warning that the new technology had been rushed into use without enough public discussion. One question is precisely how the software works -- it's the manufacturer's trade secret. Another is whether there’s scientific evidence that such technology works as advertised.
By contrast, the recent paper on the system that forecasts domestic violence lays out what it can do and how well it can do it.
One of the creators of that system, University of Pennsylvania statistician Richard Berk, said he only works with publicly available data on people who have already been arrested. The system isn’t scooping up and crunching data on ordinary citizens, he said, but is making the same forecasts that judges or police officers previously had to make when it came time to decide whether to detain or release a suspect.
He started working on crime forecasting more than a decade ago, and by 2008 had created a computerized system that beat the experts in picking which parolees were most likely to reoffend. He used a machine learning system – feeding a computer lots of different kinds of data until it discovered patterns that it could use to make predictions, which then can be tested against known data.
Machine learning doesn’t necessarily yield an algorithm that people can understand. Users know which parameters get considered but not how the machine uses them to get its answers.
In the domestic violence paper, published in February in the Journal of Empirical Legal Studies, Berk and Penn psychologist Susan Sorenson looked at data from about 100,000 cases, all occurring between 2009 and 2013. Here, too, they used a machine learning system, feeding a computer data on age, sex, zip code, age at first arrest, and a long list of possible previous charges for such things as drunk driving, animal mistreatment, and firearms crimes. They did not use race, though Berk said the system isn’t completely race blind because some inferences about race can be drawn from a person’s zip code.
The researchers used about two-thirds of the data to “train” the system, giving the machine access to the input data as well as the outcome – whether or not these people were arrested a second time for domestic violence. The other third of the data they used to test the system, giving the computer only the information that a judge could know at arraignment, and seeing how well the system predicted who would be arrested for domestic violence again.
It would be easy to reduce the number of repeat offenses to zero by simply locking up everyone accused of domestic violence, but there’s a cost to jailing people who aren’t going to be dangerous, said Berk. Currently, about half of those arrested for domestic violence are released, he said. The challenge he and Sorenson faced was to continue to release half but pick a less dangerous half. The result: About 20 percent of those released by judges were later arrested for the same crime. Of the computer’s choices, it was only 10 percent.
Berk and Sorensen are currently working with the Philadelphia police, he said, to adapt the machine learning system to predict which households are most at risk of domestic violence. Those, he said, can be targeted with extra supervision.
The parole system has already been implemented in Philadelphia. Parolees in the city are assigned to high-, medium- and low-risk groups by a machine-learning system, allowing parole officers to focus most of their attention on the high-risk cases.
One downside might be a more one-dimensional decision-making process. Several years ago, when I wrote an article on the parole system for the Philadelphia Inquirer, I learned that some parole officers found the system constraining. They said that they could have a bigger impact by spending more time with low-risk offenders who were open to accepting help in getting their lives together – getting off drugs, applying for jobs, or getting a high school degree.
Their concern was that their bosses would put too much faith in the system and too little in them. This echoes the problem Berk says worries him: That people will put too much trust in the technology. If a system hasn’t been through scientific testing, then skepticism is in order. And even those that have been shown to beat human judgment are far from perfect. Machine learning could give crime fighters a source of information in making decisions, but at this stage it would be a mistake for them to let it make the decisions for them.
http://www.bloombergview.com/articles/2016-03-08/the-crime-you-have-not-yet-committed
FBI quietly changes its privacy rules for accessing NSA data on Americans
The FBI has quietly revised its privacy rules for searching data involving Americans’ international communications that was collected by the National Security Agency, US officials have confirmed to the Guardian.The classified revisions were accepted by the secret US court that governs surveillance, during its annual recertification of the agencies’ broad surveillance powers. The new rules affect a set of powers colloquially known as Section 702, the portion of the law that authorizes the NSA’s sweeping “Prism” program to collect internet data. Section 702 falls under the Foreign Intelligence Surveillance Act (Fisa), and is a provision set to expire later this year.
A government civil liberties watchdog, the Privacy and Civil Liberties Oversight Group (PCLOB), alluded to the change in its recent overview of ongoing surveillance practices.
The watchdog confirmed in a 2014 report that the FBI is allowed direct access to the NSA’s massive collections of international emails, texts and phone calls – which often include Americans on one end of the conversation. The activists also expressed concern that the FBI’s “minimization” rules, for removing or limiting sensitive data that could identify Americans, did not reflect the bureau’s easy access to the NSA’s collected international communications.
FBI officials can search through the data, using Americans’ identifying information, for what PCLOB called “routine” queries unrelated to national security. The oversight group recommended more safeguards around “the FBI’s use and dissemination of Section 702 data in connection with non-foreign intelligence criminal matters”.
As of 2014, the FBI was not even required to make note of when it searched the metadata, which includes the “to” or “from” lines of an email. Nor does it record how many of its data searches involve Americans’ identifying details – a practice that apparently continued through 2015, based on documents released last February. The PCLOB called such searches “substantial”, since the FBI keeps NSA-collected data with the information it acquires through more traditional means, such as individualized warrants.
But the PCLOB’s new compliance report, released last month, found that the administration has submitted “revised FBI minimization procedures” that address at least some of the group’s concerns about “many” FBI agents who use NSA-gathered data.
“Changes have been implemented based on PCLOB recommendations, but we cannot comment further due to classification,” said Christopher Allen, a spokesman for the FBI.
Sharon Bradford Franklin, a spokesperson for the PCLOB, said the classification prevented her from describing the rule changes in detail, but she said they move to enhance privacy. She could not say when the rules actually changed – that, too, is classified.
“They do apply additional limits” to the FBI, Franklin said.
Timothy Barrett, a spokesman for the office of the director of national intelligence, also confirmed the change to FBI minimization rules.
Barrett also suggested that the changes may not be hidden from public view permanently.
“As we have done with the 2014 702 minimization procedures, we are considering releasing the 2015 procedures. Due to other ongoing reviews, we do not have a set date that review will be completed,” he said.
Until that hypothetical release, it remains unknown whether the FBI will now make note of when and what it queries in the NSA data. The PCLOB did not recommend greater record-keeping.
Last February, a compliance audit alluded to imminent changes to the FBI’s freedom to search the data for Americans’ identifying information.
“FBI’s minimization procedures will be updated to more clearly reflect the FBI’s standard for conducting US person queries and to require additional supervisory approval to access query results in certain circumstances,” the review stated.
Privacy advocates say that this leeway for searches that NSA and FBI officials enjoy is a “backdoor” around warrants that the law should require. In 2013, documents leaked to the Guardian by Edward Snowden revealed an internal NSA rule that Senator Ron Wyden has called the “backdoor search provision”, for instance.
While the NSA performs warrantless collection, internal rules permit the FBI to nominate surveillance targets. Those targets are supposed to be non-Americans abroad, but Americans’ data is often swept up in the surveillance.
The legal underpinnings for the dragnet, a 2008 amendment to the Foreign Intelligence Surveillance Act, are set to expire this year. A scheduled expiration of the Patriot Act last year gave critical leverage to legislators who wanted to rein in the bulk collection of domestic phone records, and intelligence officials last month implored Congress to reauthorize the measure wholesale.
“Reasonable people could and did argue about how important the telephone metadata collection was,” FBI director James Comey told the House intelligence committee last month. “This is not even a close call. This is – if we lost this tool, it would be a very bad thing for us.”
Several civil-libertarian legislators have vowed to push for an expiration of Section 702, arguing that it represents a growing surveillance authority that has moved beyond terrorism and espionage, and into the hunt for general weaknesses in the internet. The chief lawyer for the intelligence community, Robert Litt, said in 2014 that the law provides surveillance authorities the powers are “not only about terrorism, but about a wide variety of threats to our nation”.
A representative for the Fisa court deferred comment to the administration.
http://www.theguardian.com/us-news/2016/mar/08/fbi-changes-privacy-rules-accessing-nsa-prism-data
No comments:
Post a Comment