Risk Factor

This past week saw a hodgepodge of ICT-related issues. We start off with a long-standing software error affecting the credit and debit cards of some unlucky postcode related TriMet transit passengers in Portland, Oregon.

TriMet Ticketing Machine Software Error Flags Credit and Debit Cards as Fraudulent

For years, officials at Portland, Oregon’s, metro TriMet bus, light rail and commuter rail transit system have been trying to deter thieves using stolen credit and debit cards from purchasing TriMet transit tickets as a way to quickly cash in on their theft before a card is reported stolen. According to a 2011 story at the Oregonian, the thieves' modus operandi is using a stolen card to purchase an $88 TriMet pass at a ticket machine, then selling them for huge discounts in a thriving local black market. The fraud costs the transit system tens of thousands of dollars, the article says, because TriMet has made transactions using plastic so easy that “credit processor Visa requires it to cover the cost of every ticket purchased with a stolen credit card.” In 2012, Visa charged back US $95 389 for fraudulent transactions.

Many legitimate purchasers of TriMet tickets have been feeling the effects of the fraudulent activity as well. For the past several years, a large number of TriMet transit riders have been complaining that when they used their credit or debit cards to purchase a ticket, the purchases were not only declined, but their banks put security freezes on their cards out of fear that they had been stolen. Sometimes the banks would even cancel the cards outright, another story in the Oregonian reported last week.

When riders complained to TriMet about the issue, transit officials told the riders that they needed to talk to their banks about it, not them. The Oregonian stated that, “TriMet assumed problems with riders having cards suspended and cancelled were the result of banks using proprietary fraud filters to stop thieves.”

A classic case of what Oscar Wilde said about assumptions: “When you assume, you make an ass out of u and me.”

What was really behind the false positives? A software error in TriMet’s 215 ticketing machines was flagging the credit and debit cards of riders with a certain zip code as being stolen. This was happening 1000 to 2000 times a month over the past five years, Portland television station KATU reported. The error was finally discovered this January. “A data field was passing something other than TriMet's zip code, causing banks to flag the transactions as risky,” the Oregonian reported.

A TriMet official was quoted in the paper as saying, “After addressing [the error], fraud declines for credit cards users at our TVMs decreased significantly from 4 percent to 0.3 percent.”

TriMet issued a roundabout apology for the error, which was buried in a press release detailing the steps the transit agency is taking to reduce another issue angering its ridership, namely the notorious unreliability of its ticketing machines. TriMet suggests in its release that until machine reliability is improved (hopefully this summer), riders should not depend on the machines to purchase a single ticket at the station, but to instead carry a book of pre-bought tickets just in case.

Up to a Million Washington Residents Affected by Hack of State Court Network

It’s likely that most of the people charged with crimes in Washington State between September 2011 and December 2012 have already been exonerated or have paid their respective debts to society. But for roughly a million of them (at least some of whom were found not guilty at trial, established their innocence before their cases went that far—or were in court simply to fight a traffic ticket) that moment of contact with the state’s court system may lead to another punishment: identity theft. The state government revealed this week that the website for the Washington State Administrative Office of the Courts was hacked and that the attacker may have gotten away with the names and social security numbers of anyone booked into a city or county jail in the state during that time. Officials also couldn’t rule out the possibility that some people charged in the state's superior court criminal system in 2011 or 2012, cited for driving under the influence between 1989 and 2011, or went to court for traffic-related offenses during that period might be at risk. The larger group's names and driver's license information may have been taken.

"The access occurred through a ‘back door' part of a commercial software product [Adobe Systems’ ColdFusion] we were using, and it is patched now," Mike Keeling, information technology operations and maintenance manager for the court system, told reporters on a conference call.

At the same time that state officials were offering up the usual assurances that no financial data such as credit card numbers was accessed as a result of the break-in, they revealed that the breach was discovered in February (and could have been exploited as early as last fall). Since then, the state has attempted to notify only the 94 people (that is not a typo) whose information they could absolutely confirm was taken. Of their delay in reporting the incident, the government employees insisted that they didn’t initially think any confidential personal details had been stolen—despite the fact that a large volume of data had been downloaded through the backdoor. "We regret that this breach has occurred and we have taken immediate action to enhance the security of these sites," Callie T. Dietz, the state’s court administrator, said in a written statement. Dietz also offered this fun fact: The break-in was the first time the court system’s network had been hacked. Hurray! Trophies and orange slices for everyone on the team!

You may recall that I recently wrote about the apparent success of New Hampshire’s new US $90 million Medicaid Management Information System (MMIS) that went live last month after years of technical difficulties, cost overruns and delays. This was a bit of good news, given that implementations of state Medicaid/Medicare systems have a notoriously bad track record, as the project problems in Maine, Ohio and Idaho have illustrated.

Alas, the difficulty with implementing these systems was highlighted once more when late last month Tennessee announced that it was stopping work on its Vision Integration Platform (VIP) after seven years of development. According to a story in the Tennessean, the state’s Department of Human Services made only a very brief, content-free announcement about the reasons behind its termination decision on a Friday, apparently in the time-honored ploy to reduce the political impact of the news. Tennessee has had a number of high-profile IT state project problems over the past few years affecting the Department of Children’s Services, the Department of Labor and Workforce Development, as well as with the state’s attempt to implement its Project Edison payroll system.

The VIP project was to provide comprehensive automated support for Temporary Assistance for Needy Families, Food Stamps, Medicaid and TennCare, as well as other state supported programs. A February 2005 press release (pdf) from the state’s Department of Human Services said that the $37 million project would take be completed by the summer of 2008.

However, the VIP project has repeatedly missed its deadlines, with the latest being 1 April 2013. A 2012 Tennessee government audit report (pdf) blamed the missed deadlines on “defects in current designs or new functionality requirements,” the Tennessean reported. The state has spent in excess of US $20 million on the VIP project so far, and is now trying to figure out what to do next, such as to start over or to try to use what has been developed so far.

The VIP fiasco is serving to help sell Tennessee’s Governor Bill Haslam’s controversial decision announced in early April to force all of the state’s 1600 information technology workers to reapply for their jobs. Another Tennessean story says that that the purpose of the decision is to weed out “those who can’t master the skills of a rapidly changing field.”  Mark Bengel, the state’s CIO said, apparently with a straight face, “This is really not about getting rid of people. It’s about making sure that we do have the skills and we have the ability to develop and retain staff in the future.”

Last week saw the ongoing effects of several IT-related problems initially spotted over the past month. We start with problems several U.S. states have been having with the online testing systems upon which they increasingly rely for carrying out all of their standardized testing.

More States Experience Online Testing Issues

In mid-April, I wrote about how 15 000 Minnesota students who were trying to take their Minnesota Comprehensive Assessment online math test either couldn’t sign in or had their tests ended prematurely. The snag was because of a server issue at the American Institutes for Research (AIR), which the state had hired to run the testing. Those kids weren’t alone; more testing problems cropped up about a week later, frustrating the plans of several other Minnesota school districts, Minnesota Public Radio reported. However, a story at the Minneapolis StarTribune says that AIR is denying that the latter testing problems had anything to do with its system. Nevertheless, the ongoing online testing issues forced the state’s education officials to announce last week that they are pushing back the deadlines by which the exams have to be administered. The math and reading tests were to have been completed by May 10 and science tests by May 17. No new completion dates have yet to be set.

The Minnesota students having trouble taking their online tests apparently have lots of company. According to an Associated Press story, students in Oklahoma, Kentucky and Indiana also have had difficulties with their online tests. Other news reports highlighted testing issues in Ohio and Alabama. Again, server-related issues seemed to be a root cause. The problems in these states were similar to those experienced in Minnesota: students couldn’t log on, suffered slow response times, or were kicked off in the middle of their tests. Testing in Oklahoma and Indiana is run by CTB/McGraw-Hill; in Kentucky, ACT oversees the student assessments.

The problems in Indiana not only have students on edge, but teachers too, as their merit pay is tied to student test scores. McGraw-Hill, which has a four-year, $95 million contract to operate the state’s online testing system, could be fined “$50 000 for each day last week the test was down,” WTRV Channel 6 in Indianapolis reported. McGraw-Hill apologized for the problems students had in both Oklahoma and Indiana, saying “We sincerely regret the problems we have caused,” and offering assurances that everything is fine now. The apology rang pretty hollow to many teachers and parents of students affected by the outages, however.

In Kentucky, the problems have caused state officials to suspend online testing for students who haven’t completed their tests and move back to the tried-and-true, paper-and-pencil approach. According to various news reports, state officials acted after ACT gave conflicting reports on the status of its testing system.

ACT was said to have “apologized for any inconvenience" caused.

Secure Site? What Secure Site?

Researchers at security firm WhiteHat have some good news and some bad news. First the (sort of) good: for the third consecutive year, the number of serious vulnerabilities per website has gone down. But hold your applause, please. The average website was still a model of insecurity in 2012, with 56 holes.

To be sure, that's an improvement on the average of 79 per site in 2011 and an astonishing 230 per site in 2010. (By “serious vulnerability” WhiteHat means holes through which “an attacker could take control over all, or some part, of the website, compromise user accounts on the system, access sensitive data, violate compliance requirements, and possibly make headline news.”)

And now the bad news: Of the tens of thousands of sites the researchers looked at, 86 percent had least one serious vulnerability and a stunning 82 percent had a vulnerability that went unresolved for at least a full month. The most common vulnerabilities included: information leakage (55 percent of sites), cross-site scripting (53 percent), content spoofing (33 percent), and URL redirector abuse (13 percent).

Communication in a doctor's office is like a marriage gone bad: As you describe to your doctor that what pain or symptoms you have, you realize that while the doctor may hear you, he or she isn't really listening. And in the other direction, you hear the doctor's words, but do you walk away with a full understanding of the diagnosis, what exactly you're being prescribed, why, and what the risks are?

In a recent article in the London Telegraph, fully 25 percent of National Health Service patients complained that their doctors discuss their conditions as if they weren’t there; 20 percent reported that “they were not given enough information about their condition and treatment;” and 25 percent confessed that “there was no one they could talk to about their worries and fears.”

Another recent story about doctors’ people-skills—and lack thereof—in the Wall Street Journal sums up the issue nicely: “Doctors are rude. Doctors don't listen. Doctors have no time. Doctors don't explain things in terms patients can understand.” The introduction of electronic health records has, ironically, often made things worse. A recent study noted that even as EHR systems have "allowed [doctors] to spend more face-to-face time with patients," they nonetheless often prove to be a "distraction" as doctor attention becomes focused on keyboards and not patients.

The WSJ article talks about how medical schools, malpractice insurers, and major hospitals are trying to improve patient-doctor communication, and for good reason: A break-down in patient-doctor communication is cited in at least 40 percent of malpractice claims. Further, research confirms that poor communication often leads patients to not follow their prescribed treatments regimens, whereas the opposite also seems to be true. Doctor-patient communication can be improved by having doctors coached in practices like the Four Habits, which, the WSJ says, “teaches doctors how to create rapport with patients, elicit their views, demonstrate empathy and assess their ability to follow a treatment regimen.”

If new technology is partly to blame, it can also help. An article at Health Management Technology describes the use of speech-based “virtual assistants” to capture patient data and automatically enter it into a patient’s EHR, for example, allowing the doctor to talk to the patient without the distraction of having to type what is being said.

Deal of the Week: Identity Theft

On Saturday, I and 50 million other LivingSocial customers received e-mail notices from company CEO Tim O'Shaughnessy telling us that we got more than we bargained for when we signed up to receive deals via the site. We got hacked.

Earlier in the week, our account details including names, e-mail addresses, birth dates, and encrypted passwords had been compromised. The e-mail told us that LivingSocial had already reset users’ passwords, mainly to force customers to create new ones. The note assured us that our credit card information was stored on a separate server that was not breached and thus did not fall into the attackers’ hands. Though LivingSocial also offered assurances that our stored passwords were encrypted, security experts laugh at the notion that a highly motivated hacker will be stymied by that barrier. What will the cyberthieves do with the information? PC World notes that they're not likely to run up a tab of discounted facials, massages or walking tours. “The bigger concern is what an attacker can do with your personal information,” the article notes. It warns customers to change their passwords on other sites if they made the all-too-common mistake of using a single password repeatedly. For its part, LivingSocial told us that “The security of your information is our priority. We always strive to ensure the security of our customer information, and we are redoubling efforts to prevent any issues in the future.” Have you ever noticed that no CEO ever sends out an e-mail that says, "We've never been hacked, but we're redoubling efforts anyway"?

Hacking the Meaning of Hacking

It’s happened before: someone is convicted for robbery who never set foot inside the store that was held up, or serves a long prison stretch for murder, but is later exonerated when DNA evidence reveals that they and the perpetrator are not one and the same. But you rarely associate computer crimes with such miscarriages of justice. Nevertheless, in a California courtroom this week, David Nosal was convicted of six counts, including violating the federal Computer Fraud and Abuse Act—which went on the books in 1984 as part of an effort to make it easier for prosecutors to take down hackers bent on stealing data or in some way vandalizing the machines they infiltrate. The problem: There is no question about the fact that he did not hack into the system from which he acquired proprietary information. The jury came back with a guilty verdict despite having heard evidence that Nosal managed to convince—mostly through bribery—his former colleagues who were still employed at Korn/Ferry International, an executive search firm, to access the company’s database and turn over trade secrets. And get this: Those folks, who actually accessed the Korn/Ferry database with malicious intent, were not charged with any criminal wrongdoing.

But Nosal likely won’t don prison stripes anytime soon. If the pattern of this case holds, the verdict is, for the accused, merely a setback in a long and winding journey. The judges of the Ninth Circuit Court of Appeals in San Francisco have banged their gavels on this case on two separate occasions, and legal observers say they’re likely to see it again. Last year, the Ninth Circuit jurists decided that bringing charges against an employee for what amounts to a violation of his or her employer’s computer use policy is a bridge too far. That saved the bacon of Nosal’s aforementioned accomplices and got some charges against him related to data thefts back when he was a still a Korn/Ferry employee dropped. Furthermore, chances are good that a final decision on Nosal’s fate won’t be made until the Supreme Court weighs in. Stay tuned.

I am only surprised it took so long.

Yesterday, a “breaking news” tweet at 1:07 PM EDT from the Associated Press reported that two explosions had occurred at the White House and President Obama had been injured. The news immediately sent the Dow Jones Industrial Average down 143 points, as this graph at the London Telegraph shows. There's also a lovely animated display of the “flash crash” by market research firm Nanex LLC.

It took about three minutes for the tweet to be repudiated, and a bit longer for the AP to acknowledge that its Twitter account had indeed been compromised. According to its own story posted last night, all of the AP's Twitter accounts (including its Mobile Twitter account) were suspended and it was “working to correct the issue.” The AP also stated that the “Syrian Electronic Army claimed responsibility for the hack,” but added that, “This couldn’t be corroborated.”

The SEA, which supports the Syrian Government, has taken credit for a number of recent Twitter account compromises, including the  BBC, National Public Radio, CBS News and the President of FIFA. Last August, there were a number of fake news stories published regarding the Syrian conflict as well. Facebook is also a Syrian conflict social battleground.

A more intriguing statement in the AP story generated lot of speculation: “The attack on AP's Twitter account and the AP Mobile Twitter account was preceded by phishing attempts on AP's corporate network.” This suggests that someone in the AP downloaded a phish email (seemingly confirmed by AP reporter Mike Baker) which led to the compromise of the AP Twitter accounts. However, when asked for clarification, the AP refused any further comment, maybe on the advice of the FBI and the U.S. Securities and Exchange Commission, who are looking into the incident.

Recent comments by the Boston Police Department to the Washington Post confirm what Tech Talk said last week: Facial recognition was not a factor in the hunt for the Boston Marathon bombers.  According to the Post, “facial-recognition software did not identify the men in the ball caps. The technology came up empty even though both Tsarnaevs’ images exist in official databases: Dzhokhar had a Massachusetts driver’s license; the brothers had legally immigrated; and Tamerlan had been the subject of some FBI investigation.” Image analysis software has not caught up with the grimy reality of street photography: low-resolution, long-range images—often poorly focused, rapidly moving, and caught from odd angles.  

But if analytical cameras were not a factor last week, they may be soon.

The New York Police Department and Microsoft have built what they call the “Domain Awareness System.”  Early in April, before the Boston bombing, the New York Times described the combination of “more than 3500 cameras in public places, license-plate readers at every major Manhattan entry point, fixed and portable radiation detectors, real-time alerts transmitted from the 911 emergency system and a trove of Police Department data, including arrests and parking summonses.”

The system can spot cars on its watch list the moment they enter the city, and trail them by camera throughout their trip. The system also uses medium-time-scale scene analysis to spot and report changes in the relatively static features of a video image—a backpack left by a railing on a busy street, for example—while ignoring rapidly changing components like people and vehicles passing by. It’s like those spooky, ultra-long-exposure pinhole camera photos that turn a busy street into an empty post-apocalyptic dreamscape, unpopulated under the noonday sun. You can see demonstrations of the Domain Awareness System in action in this Today Show segment. (The technique has broad application to cluttered environments. See the LandTrendr analysis that researchers from Oregon State University and the U.S. Department of Agriculture used to transform Landsat satellite images of the Pacific Northwest into a pristine, cloud-free, season-by-season time series yielding a clear picture of bark beetle damage creeping across the region over 23 years.)  

Why is the NYPD going public with a description of its high-tech crime-fighting tools? New York Police Commissioner Ray Kelly says that knowing the odds against them will deter potential terrorists. We should also note that the NYPD and Microsoft have recently agreed to market the system, and New York stands to receive 30 percent of the gross revenues on the multi-million-dollar licenses.