Researchers from the University of California San Diego, Purdue University, UNC Chapel Hill, Georgia Institute of Technology, and Google have uncovered two (2) new attack methods targeting high-performance Intel CPUs, collectively named “Pathfinder”. These methods exploit vulnerabilities in the branch predictor, particularly the Path History Register (PHR), to launch attacks that can reconstruct a program’s control flow history or execute high-resolution Spectre attacks. Spectre attacks utilize branch prediction and speculative execution to bypass application isolation protections and access privileged data. The new techniques enable attackers to manipulate the PHR and the prediction history tables, leaking historical execution data and potentially allowing them to recover AES encryption keys or extract secret images processed by libraries such as libjpeg. Intel has responded by noting that existing mitigations for previous Spectre attacks also mitigate these new exploits. The study was responsibly disclosed to Intel in November 2023, and the findings highlight significant vulnerabilities in modern CPU designs that are not present in AMD CPUs. CTIX analysts will continue to report on new and evolving forms of malware and associated campaigns.