Personal RFID devices found, e.g., in access cards and contactless credit cards are vulnerable to unauthorized reading, owner tracking and different types of relay attacks. We observe that accessing a personal RFID device fundamentally requires moving it in some manner (e.g., swiping an RFID access card in front of a reader). Determining whether or not the device is in motion can therefore provide enhanced security and privacy; the device will respond only when it is in motion, instead of doing so promiscuously. We investigate extending the concept of min-entropy from the realm of random number generation to achieve motion detection on an RFID device equipped with an accelerometer. Our approach is quite simple and well-suited for use on low-cost devices because the min-entropy of an accelerometer's distribution can be efficiently approximated. As opposed to alternative methods, our approach does not require any changes to the usage model expected of personal RFID devices.