Researchers hack 'tamper-proof' PIN terminals

Researchers hack 'tamper-proof' PIN terminals

Summary: Three Cambridge University researchers have succeeded in hacking two of the most widely used PIN terminals in the UK, using a low-tech combination of paperclips and FPGA boards

TOPICS: Security

Researchers from Cambridge University have succeeded in capturing both PIN numbers and card details from supposedly tamper-proof PIN terminals.

Saar Drimer and Steven Murdoch, overseen by Professor Ross Anderson, managed to hack two widely used PIN terminals: the Ingenico i3300 and the Dione Xtreme.

In a research paper seen by, the researchers outline the hack. Both terminals have tamper-proof mechanisms inside, but both can be circumvented by tapping the data line of the PIN Entry Device/smartcard interface. The data exchanged on this line is not encrypted.

The Ingenico i3300 has a tamper-response switch inside which is tripped if the terminal is forced open, and also has its innards wrapped in a tamper-proof mesh, to detect drilling. However, there is a user-accessible compartment to insert SIM cards that is not intended to be tamper-proof. The PCB has various holes that an attacker can use to insert a conductor into the serial data line, to tap both the PIN and card details. The researchers used a paper clip as the conductor, linked to the data line.

The Dione Xtreme also has a tamper-response switch, but no mechanisms to detect drilling from the rear. The main keypad and processor are "potted together", making it more difficult to incept the signal passing between them. However, by drilling a 0.8mm hole from the rear, the researchers inserted a 4cm needle into a flat ribbon connector socket and tapped the data.

In both cases, the conductors were connected to a thin wire connected to a logic board containing a field programmable gate array (FPGA), which translated the data and sent it to a laptop.

Both devices were Visa-certified to be secure, which requires that defeating the tamper detection would cost over $25,000 (£12,500) per-PIN entry device; or that inserting a PIN-stealing bug would be detected, or take more than 10 hours.

Neither terminal meets any of these requirements, said the researcher paper.

"What should have required $25,000 needed just a bent paperclip, a needle, a short length of wire and some creative thinking; attaching them to the data line takes minutes with some practice," said the paper.

"What this shows is that PIN entry devices in the UK are very insecure," said Professor Anderson about the research. "What's more, the [device] certification process is completely defective. Certified devices are easy to breach. That's bad news for retailers, and bad news for customers."

Read this


Special report: Anatomy of a hack attack

We recreate a typical attack on two large organisations

Read more

Drimer added that this hack showed the complete process from design to implementation of these devices was broken.

"These devices should not have been certified because they clearly fail the criteria under which they were evaluated," Drimer told "Something went wrong in the design of these devices, the certification process, and the EMV implementation choices made by the banks."

Ingenico admitted that the hack was successful, but said that its device "still remained one of the safest on the market".

A spokesman for Ingenico Northern Europe said: "Retailers and card users should rest assured that the devices, from various suppliers, identified by the Cambridge University scientists, remain among the most secure terminals on the market and have contributed to card fraud at UK retailers falling by up to 47 percent year-on-year since the introduction of chip and PIN. The banking industry has already expressed its confidence in the security capabilities of all chip and PIN payment devices being used in the UK today.

"The method identified by the Cambridge University paper requires specialist knowledge and has inherent technical difficulties. This method is therefore not reproducible on a large scale, nor does it take into account the fraud monitoring used throughout the industry."

"Security remains a top priority for Ingenico and we invest around €40m each year in research and development to ensure our customers remain at the forefront of the fight against fraud."

"This investment is highlighted in the latest generation of our terminals which are approved under the latest security standards. These meet the higher security required by industry mandates introduced on 1 January, 2008 and are designed to stay one step ahead of the evolving security threat."

Dione, which is manufactured by Verifone, had not responded to a request for comment at the time of writing.

Topic: Security

Tom Espiner

About Tom Espiner

Tom is a technology reporter for He covers the security beat, writing about everything from hacking and cybercrime to threats and mitigation. He also focuses on open source and emerging technologies, all the while trying to cut through greenwash.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


1 comment
Log in or register to join the discussion
  • drilling into the back of an autoteller

    I'm not paid to do security in any way, as an amateur and customer of a bank, my thoughts:

    Drilling into the back of an autoteller is detectable often I believe:

    The back of the autoteller is either in a building or outside:
    Outside they are normally clad in concrete and possibly also reinforced. Heavier drilling equipment, and more noise are more detectable. Also it's more difficult to be precise where the drill bit goes (necessary to avoid the antitamper measures)

    If the back of the autoteller is inside a building it may be less well physically protected but there is the security of the building; daytime, the human supervision, and at night whatever alarms detectors &c the banks or shops use to protect the building. A break-in is usually detected.

    I think the manufacturers sound complacent. And I don't buy their line that you need to be a cambridge boffin to make the hack work.
    It sounds like security by obscurity. Skilled electronics engineers are available for hire and not all are squeaky clean.

    I speculate the hack used an FPGA for cheapness, and to demonstrate reproducibility, not for speed of processing.

    Would the researchers care to comment?