After uncovering a security weakness in a radio-frequency identification tag from Texas Instruments Inc., researchers from RSA Security Inc.s RSA Laboratories and The Johns Hopkins University are now eyeing future exploits against other RFID products in the interests of better security, one of the researchers said this week.
Meanwhile, TI will keep making the compromised RFID tag in order to meet the needs of applications more sensitive to speed and pricing than to privacy, according to a TI official.
The Johns Hopkins University Information Security Institute and RSA first publicized their findings about the RFID security hole in January.
In a paper posted at www.rfidanalysis.org, the researchers claim that by cracking a proprietary cipher, or encryption algorithm in one of TIs DST (digital signature transponder) RFID tags, they were able to circumvent the tags built-in security enough to buy gasoline and turn on a cars ignition.
Tony Sabetti, global business manager for TIs RFID Systems, acknowledged that the DSTs contain some proven vulnerabilities. But Sabetti also described the security risk as relatively minimal, calling it a “tradeoff” that some makers of electronic payment and vehicle immobilization systems are willing to accept. Some of TIs customers in these niches produce car keys or tokens, and others, complete systems.
The RFID tags compromised by Johns Hopkins and RSA—part of TIs DST-40 tag lineup—use a proprietary 40-bit encryption algorithm first written in 1999.
“Why are we using a proprietary algorithm? Because its faster [that way] to produce inexpensive chips,” Sabetti said.
The researchers from Johns Hopkins and RSA reverse-engineered and emulated the 40-bit encryption over two months.
But DST-40 tags are only one part of a larger RFID portfolio that also includes a DST “Plus” edition—featuring “a series of memory features and encryption scalable to 80 bits”—as well as an “RFID credit card” lineup with industry-standard 128-bit Triple DES encryption, he said.
TI keeps “vigilantly improving” products across this portfolio, according to Sabetti. But, he said, TI has no immediate plans to stop using the proprietary 40-bit cipher in its DST-40 tags. Customers that choose DST-40 tags from TIs lineup are generally seeking a combination of low pricing and quick processing speeds.
In a vehicle immobilization application, for example, it takes the 40-bit encryption scheme only about 250 milliseconds to “wake up the tag, do the encryption and encoding, and confirm that everythings correct,” Sabetti said.
“But if you put [128-bit] Triple DES in there, all this would take 2 to 3 seconds—and that wouldnt be acceptable to most consumers.”
Few people would be able to replicate the attack, anyway, Sabetti said. “Deciphering a 40-bit key isnt really that special of an event. But this was the work of professional mathematicians, statisticians, and cryptographers.”
Moreover, TIs customers add other levels of security at the application layer, according to Sabetti.
Next page: Striking a balance between speed, cost and security.
Page Two
One of TIs customers, Exxon Mobil Corp., uses the DST-40 tag in its Speedpass electronic payment system.
“Weve been over all of this with Exxon, and they dont see any more risk [now] beyond the risk they were at already,” Sabetti said.
Ari Juels, RSAs principal research scientist, agreed that customers have a right to try to “strike a balance” between speed, cost, and security.
“But what were saying is that if youre going to bother to build in encryption at all, you should do it correctly. You shouldnt use a proprietary encryption algorithm,” Juels said.
Other companies also make RFID tags for the same general markets. Why did the researchers focus only on TI? “Because TI is the most visible.” Juels said.
But TI wasnt being “picked on,” he said. “Were also intending to look at a range of [other] RFID systems, [including both] active and passive RFID tags. Various [RFID systems] have various security and privacy weaknesses.”
For his part, Sabetti took issue with some of the cryptographic researchers methods and findings.
The DST simulation system used in the attacks “took up the entire back seat of a car,” he said. “[The researchers] were unable to produce an emulator which would be considered small or efficient.”
Sabetti also said that the RSA and Johns Hopkins researchers had demonstrated the attack for him, and that during the demo, they hadnt been able to intercept information outside of a two-foot range.
“[Two feet] is a bit naïve,” Juels said. “That is the nominal read range. Other systems, [with] a gate antenna, might achieve several feet for active scanning. Also, if an attacker waits until someone is [actually] using a Speedpass token, the potential passive eavesdropping range could be 10 feet.”
But Juels admitted that the researchers used only “crude” equipment in the attack against TIs 40-bit encryption.
“Our attempt to do this was rather crude and uninformed, [and] cobbled together with some fairly inexpensive equipment. [But] once these systems have been widely deployed, there will be better equipment available. This was only a proof-of-concept [attack],” he said.