Tag: side-channel attacks

  • The Register: Boffins trick AI model into giving up its secrets

    Source URL: https://www.theregister.com/2024/12/18/ai_model_reveal_itself/ Source: The Register Title: Boffins trick AI model into giving up its secrets Feedly Summary: All it took to make an Google Edge TPU give up model hyperparameters was specific hardware, a novel attack technique … and several days Computer scientists from North Carolina State University have devised a way to copy…

  • Hacker News: Computing with Time: Microarchitectural Weird Machines

    Source URL: https://cacm.acm.org/research-highlights/computing-with-time-microarchitectural-weird-machines/ Source: Hacker News Title: Computing with Time: Microarchitectural Weird Machines Feedly Summary: Comments AI Summary and Description: Yes Summary: The text discusses the development and implications of microarchitectural weird machines (µWMs), which exploit CPU microarchitectural features to create powerful obfuscation techniques for malware. This research provides insights into how these µWMs can…

  • Hacker News: When machine learning tells the wrong story

    Source URL: https://jackcook.com/2024/11/09/bigger-fish.html Source: Hacker News Title: When machine learning tells the wrong story Feedly Summary: Comments AI Summary and Description: Yes Summary: The text presents a comprehensive account of a research project that explores the vulnerabilities of machine learning-assisted side-channel attacks, particularly in web browsers. It emphasizes the implications of using machine learning for…

  • Hacker News: Cross-Process Spectre Exploitation

    Source URL: https://grsecurity.net/cross_process_spectre_exploitation Source: Hacker News Title: Cross-Process Spectre Exploitation Feedly Summary: Comments AI Summary and Description: Yes **Summary**: This detailed text discusses a newly developed cross-process Spectre attack exploiting vulnerabilities in Intel processors, specifically involving the Indirect Branch Prediction Barrier (IBPB). The attack showcases how certain exploit methodologies can retain unauthorized data through microcode…