《Immersive Lab：2024生成式AI（GenAI）的阴暗面-人们如何诱使聊天机器人泄露公司机密研究报告（英文版）（26页）.pdf》由会员分享，可在线阅读，更多相关《Immersive Lab：2024生成式AI（GenAI）的阴暗面-人们如何诱使聊天机器人泄露公司机密研究报告（英文版）（26页）.pdf（26页珍藏版）》请在三个皮匠报告上搜索。

1、Unveiling the Dark Side of GenAI:How People Trick Bots into Revealing Company SecretsDr.John Blythe,Director of Cyber Psychology,Immersive LabsKev Breen,Senior Director of Threat Intelligence,Immersive LabsJoel Iqbal,Data Analyst,Immersive LabsContentsCopyright 2024 Immersive Labs.All rights reserve

2、d.02Unveiling the Dark Side of GenAI|May 2024Introduction3Key Takeaways6Methodology7Findings11Why Organizations are at Risk12The 12 Most Commonly Used Prompt Techniques13User Techniques by Level15How People Use Role Play to Gain an Advantage16The Length and Number of Prompts Used17The Sentiment of U

3、ser Prompts20GenAI Risk:A Psychologists Perspective23Conclusion24Urgent Calls to Action26Copyright 2024 Immersive Labs.All rights reserved.03Unveiling the Dark Side of GenAI|May 2024IntroductionCopyright 2024 Immersive Labs.All rights reserved.04Unveiling the Dark Side of GenAI|May 2024“Artificial i

4、ntelligence(AI)will almost certainly increase the volume and heighten the impact of cyber attacks over the next two years.”1UK National Cyber Security Centre 2023?https:/ Artificial Intelligence(GenAI)has emerged as a transformative technology,revolutionizing various aspects of our lives.With its ab

5、ility to mimic human intelligence and perform complex tasks,GenAI has gained immense popularity across industries and among users worldwide.But with the prevalence of these tools comes novel cybersecurity risks.GenAI-powered Large Language Models(LLMs)like OpenAIs ChatGPT and Googles Gemini models r

6、ely on user prompts to generate responses.While LLMs are highly adaptable to different tasks and applications,they can be manipulated by people with potentially damaging consequences,emphasizing the need for careful consideration of the various security implications.LLMs expose organizations to prom