美国国防部计划在5年内投入3030万美元开发代号"Polygraph+"的项目，以人工智能和机器学习技术改进传统测谎手段1。该项目由防御反情报和安全局执行，主要用于员工审查和内部威胁检测1。
新系统采用多项创新技术，包括远距离生理传感等无需直接接触受测者的检测方法1。2023年国防创新部已选中两家企业参与开发：Presage Technologies利用标准摄像头测量心率和呼吸，Altec Research则追踪头部运动、面部皮肤温度和毛孔活动1。
不过，学术界对这一方案持怀疑态度。Kyri Kotsoglou教授称这是"一个误导性努力，试图将复杂问题简化为有形的东西"1。Marion Oswald教授则担忧，新型测谎技术可能被用作心理威胁工具而非科学工具，且很可能是对当前行政部门泄密担忧的回应1。从效能角度看，美国国家研究委员会在2003年的评估表明测谎仪的有效性证据"充其量较弱"1。美国多边形协会虽声称准确率达80-94%，但若应用于280万国防部员工，仍可能导致数万人被错误指控1。
The U.S. Department of Defense is requesting $30.3 million over five years to develop an artificial intelligence-enhanced polygraph system called "Polygraph+" or "Polygraph Next," aimed at improving employee vetting and internal threat detection 1. The Defense Counterintelligence and Security Agency (DCSA) will oversee the project, which combines machine learning algorithms with novel contactless biometric sensing technologies 1.
The initiative draws on technology already tested by the Department of Defense Innovation Unit, which selected two companies in 2023: Presage Technologies, which measures heart rate and respiration using standard cameras, and Altec Research, which tracks head movement, facial skin temperature, and pore activity 1. However, academic experts have expressed significant skepticism about the approach. Professor Kyri Kotsoglou characterized it as "a misleading effort that attempts to reduce complex problems into something tangible," while Professor Marion Oswald warned that such technology could function as a psychological intimidation tool rather than a scientific instrument, potentially representing a response to current administrative concerns about leaks 1.
The scientific foundation for the project remains contested. The National Research Council concluded in 2003 that evidence for polygraph validity was "at best weak" 1. While the American Polygraph Association claims accuracy rates between 80 and 94 percent, applying such technology to the Pentagon's 2.8 million employees could result in tens of thousands of false accusations if error rates fall within typical ranges 1.
评论
还没有评论，欢迎留下第一条。