This paper examines the operation of TFRC (TCP-Friendly Rate Control) in scenarios where the receiver is untrustworthy. By misbehaving, a TFRC receiver can obtain a higher data rate at the expense of competing traffic. We identify and experimentally confirm several such attacks and designed Robust TCP-Friendly Rate Control (RTFRC), a TFRC variant which is resilient to receiver misbehavior. We also show that additional attacks that are based on feedback timing and targeted directly at RTFRC are unable to compromise the protocol. We discuss existing and propose new techniques for protecting congestion control protocols from receiver misbehavior in general. The discussion includes analysis of what level of protection is feasible with different amounts of feedback. Finally, we explore methods that compress feedback reports without undermining their verifiability.