Your household appliances will soon be talking together over the Internet, but what will they say about you? Will they give away your secrets?
We are moving toward a world where more than just your car, your phone and your computer will be linked over the Internet. Home improvement stores are planning home networks where many basic functions of your home, including the heat, the water, the security system and the major appliances can be accessed, activated and updated through the web. Although this technology is in its nascent stages, we are already seeing security problems.
The Federal Trade Commission has made the regulation of the so-called "Internet of Things" a top priority for the agency. The FTC describes the "Internet of Things" as an "everyday product with interconnectivity to the Internet and other mobile devices." In particular, FTC officials say they will strictly regulate and enforce consumer privacy concerns.
Recent FTC actions show that regulators aren't just making idle threats. A home-security camera maker has been cited by the Federal Trade Commission for poor security and false security promises, when the cameras could easily be hijacked over the Internet, allowing voyeurs and plotters to spy inside the houses and businesses of camera-using consumers. The action against TRENDnet marks the FTC's first action against a marketer of so-called "Internet of Things" products
In September, the FTC announced a proposed settlement with TRENDnet, a California-based company that markets routers, modems and Internet Protocol cameras to home users and businesses to monitor their homes or businesses from a computer, smartphone, tablet or other mobile device. The FTC's complaint alleges that TRENDnet failed to implement reasonable security measures, and as a result the live feeds for nearly 700 cameras were available to the public on the Internet. The live feeds displayed private areas of the users' homes, including images of sleeping infants, young children playing and other daily activities of consumers. One hacker posted information about the breach online, spreading these live video feeds across the Internet.
In the FTC complaint, the agency says TRENDnet's lack of security caused significant injury to consumers who bought and used the cameras.
"The exposure of sensitive information through respondent's IP cameras increases the likelihood that consumers or their property will be targeted for theft or other criminal activity, increases the likelihood that consumers' personal activities and conversations or those of their family members, including young children, will be observed and recorded by strangers over the Internet. This risk impairs consumers' peaceful enjoyment of their homes, increases consumers' susceptibility to physical tracking or stalking, and reduces consumers' ability to control the dissemination of personal or proprietary information," the complaint states.
According to the FTC's complaint, TRENDnet marketed its products as "secure" and used the trade name "SecurView." Packaging and other promotional material emphasized that the products would help customers "protect" their homes and property. But the FTC claims that TRENDnet failed to use reasonable security measures by engaging in the following practices:
- Transmitting and storing user login credentials over the Internet and on mobile devices in clear, readable text, when readily available software exists to secure such transactions;
- Failing to implement a process to actively monitor security vulnerability reports from third-parties;
- Failing to employ reasonable security in the design and testing of its software; and
- Offering a faulty setting that purported to give users the option to require login credentials to access the camera.
Under the proposed terms of the FTC settlement, TRENDnet is prohibited from misrepresenting the security of its products and the information that its products transmit, and the extent to which a consumer can control the security of any information the devices store or transmit.
In addition, the company must notify affected consumers and provide free technical support to consumers over the phone and email for the next two years. TRENDnet must also create a comprehensive security program and undergo third-party audits for the next 20 years.
TRENDnet officials say the breach has been a learning experience and they are taking steps to ensure that such a compromise of consumer privacy will not happen again.
The FTC has made an important strategic practice of attacking companies for lax security by claiming that such behavior violates representations of privacy and security to consumers. Then the FTC insists on reaching a 20 year consent order with the target company, which it uses to more tightly regulate the company's privacy practices in a manner that the FTC Act might not otherwise support. The FTC has done this in the social media space with Facebook, MySpace, Twitter and Google, and it seems to be porting this strategy to companies providing connectivity to the everyday things in our lives.
Concerned companies should provide strong data security, including written policies and procedures that can be defended in court and that grow with the changing state of the art. Also be sure that your company's published policies match, but do not exceed, your company's actions and standards in protecting consumer data.
A Billion Machines Making Decisions For Themselves: What Could Possibly Go Wrong?
As the Internet of Things becomes pervasive, companies and societies will need rules for the new sensor-driven business and operational models. This is a place where lawyers are needed. When a self-driving car makes a wrong turn into a swimming pool, who is at fault? Does the driver maintain all responsibility? What if there is no driver? Does the programmer of equipment manufacturer accept any of the legal blame? What if the vehicle was fooled by torrential rain or other extremely rare condition? Does this absolve the driver, the programmer, the manufacturer, or anyone else in the chain of responsibility?
And when machines contract with each other, what happens when one side breaks the contract? How do you measure a "meeting of the minds" when no minds were involved in the transaction? We are already seeing computers make split-second buying and selling decisions in the securities and commodities markets, so we are likely to have rules worked out for machine-to-machine contracting in this space, and those rules are likely to spread to other fields by analogy. And where liability decisions live, insurance is sure to follow. Insurers will need to build models to decide when to accept liability for the lousy decisions of its customers' automated devices. It is likely that machine decisions will be excluded from policies first, so that the industry can begin to calculate how to charge for insuring these risks.
Privacy Violations
Other risks in this space include the loss of personal privacy suffered by individuals whose homes, trucks, offices and bodies are constantly monitored and analyzed by everything from pulse-sensing watches to exercise machines provided by insurance companies to monitor health improvement efforts. From cars that know when you break the law (currently existing) to houses that tell you which part of the day to operate your washing machine (coming soon), your devices will collect more information about your lifestyle and activities than you thought possible, and they will be sending this information over the Internet to manufacturers and other interested parties. Sometimes worse than the businesses collecting your data, are the bad guys hacking into either the business information or to the devices themselves to gain an insight into your life.
The FTC's action reported above is the first of many acknowledgements that companies peering into our homes are not always careful and protecting the pictures or the data that they collect. Now add the concern that the sensors themselves – like the cameras in the FTC case – may be controlled by someone with no authorization and a malevolent motive. In this case, the malevolent party can see inside your home or car, and can take vital information to rob you or worse. The deeper that sensors burrow into your life, the more intimate information can be extracted and used. Some companies (and political parties) are building profiles with this data, so that they know how reach (and manipulate) you easier.
Hacking Your Stuff and Taking Control
The true malevolence comes in taking over Internet-connected devices that have an important function, rather than just sensors that provide information. Rather than just seeing into your life, hackers can use these physical machines to invade your life and make the things around you malfunction. When mechanical devices are controllable online, then that control can fall into the wrong hands, allowing hackers to shut off your refrigerator, spoiling the food, or leave your hot tub running all weekend, thus running up your electric bills. However, the truly frightening scenarios involve more dangerous circumstances, for example, affecting the breaks or power steering in your car or turning your insulin pump into high gear.
Remote Control Vehicles
For example, in the most recent Black Hat Asia security conference, Spanish researchers presented a device that can penetrate the network of an automobile and sabotage its functions. Many people do not know that their cars and trucks are controlled by individual computer networks contained within the vehicle, but in the past decade automobiles have benefited from the various revolutions – storage, miniaturization, digitization, connectivity – that are leading us to the Internet of Things. If you buy a new vehicle today, it will contain a black box – just like an airplane – to record the vital information from all of the vehicle's sensors in the minutes before a crash. But more than sensors are connected to the computer.
Everything electric, from the windows to the wipers to the brakes are controlled, in whole or in part, by an onboard network called the Controller Area Network bus (the CAN bus) that coordinates and operates these features. The vehicle intrusion device discussed by the Spanish researchers is smaller than a smartphone and, when hacking into your car, it draws power from the vehicle's electrical systems and connects to the CAN bus to input commends from a remote attack computer. This device, known as the deviCAN hacking tool, must be attached to the car to work, and it connects into the CAN bus with four small wires, but once attached, it can allow a remote attack computer to control or shut down any of the automotive systems connected to the CAN bus. While any device that contains or is controlled by a processor is hackable, engineers have not normally thought of embedded device systems as vulnerable to outside electrical hijacking. While the deviCAN hacking tool is not the first to take advantage of this idea, it is inexpensive and easy to build with accessible parts.
Hacking Your Heart
Computer hacker and security company executive Barnaby Jack, who died of an apparent drug overdose in 2013, was best known for showing a security conference audience the easiest methods for "jackpotting" ATM machines – hacking the machine so it will spew money forth without debiting the funds against any particular account. However, the week after he died, Jack was scheduled to give a talk to another security conference about the hacking of medical devices. Jack gave an interview about the subject two weeks earlier where he described his discovered method of remotely disabling a pacemaker, leading to the interview headline in Vice magazine: "BARNABY JACK COULD HACK YOUR PACEMAKER AND MAKE YOUR HEART EXPLODE." When asked about the difficulty of hacking into embedded medical devices like pacemakers and insulin pumps, Jack said, "It does take a specialized skill, but with more and more security researchers concentrating on embedded devices, the skill set required is becoming more common. It probably took me around six months, from reverse engineering and finding the flaws through to developing software to exploit the vulnerabilities." His pacemaker hack included providing a profound electrical shock to the device's host. Security researcher Jerome Radcliffe demonstrated and the 2011 Black Hat conference that he could wireless hack into his own insulin pump and manipulate the amount of insulin administered.
Any device that is networked could also be hacked, and there are important reasons that device manufacturers leave connectivity openings into these embedded devices, including the ability to fix problems remotely if the device begins to malfunction within a person's body. It is not just manufacturers who put their own devices at risk, as security implementation by doctors and hospitals can also be an important area where security practices fall short. The networks hosting these devices can be misconfigured, providing exploitable vulnerabilities, and bug fixes are not always immediately released or applied because medical offices are not used to applying the security protocols that computer technologists in other businesses have formed as habit for years. Hospitals often use out of date software and their networks can be filled with malware.
The Center for Internet Security, a nonprofit group that advises companies and government agencies, is working on a set of guidelines for securing imbedded medical devices. The U.S. Food and Drug Administration is developing its own guidelines in this space, finding that "Addressing cybersecurity threats and reducing information security risks is especially challenging because of the need to balance the protection of patient safety with promoting the development of innovative technologies and improved device performance." FDS Statement on Medical Devices, November 1, 2013. However, as of this date, the FDA has not produced anything more specific than highlighting the need to be vigilant about cyber hacking issues in the medical device market. In June of 2013, ICS, part of the Department of Homeland Security, reported that it had found security holes in 300 medical devices being made by 40 different companies. It is not only embedded devices that are problematic, but any connected surgical device, like breathing machines, can kill a patient on the table, and dozens of other life-sustaining devices in hospitals can be hijacked to the same end. The types of machines studied by ICS including Surgical and anesthesia devices, ventilators, drug infusion pumps, external defibrillators, patient monitors and laboratory and analysis equipment. The latter category of machines could be used to provide mistaken readings to doctors, leading to the application of dangerous treatments, or mistaken non-treatment of cancers or other diseases.
In February of this year, it was discovered that hackers had penetrated the computer networks of the nation's top medical device makers, including Medtronic, Boston Scientific and St. Jude Medical. The companies were under federally imposed obligations to disclose these breaches, but did not do so. While these hacks seemed to come from China and apparently were aimed at device manufacture data and other intellectual property, similar hacks could be used to find the basic schematics and security functionality for medical devices, making the devices easier to hack and turn against the patient. Vulnerabilities abound in the medical system, and the introduction of a much more robust Internet of Medical Things will only exacerbate the situation.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
