From dc3229b2a5a77ec97014113e5a10fa09f258b12a Mon Sep 17 00:00:00 2001 From: shiva108 Date: Sat, 10 Jan 2026 14:13:39 +0100 Subject: [PATCH] Small reading friendly fixes to 24 --- docs/Chapter_24_Social_Engineering_LLMs.md | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-) diff --git a/docs/Chapter_24_Social_Engineering_LLMs.md b/docs/Chapter_24_Social_Engineering_LLMs.md index 34c63f4..fc5ab78 100644 --- a/docs/Chapter_24_Social_Engineering_LLMs.md +++ b/docs/Chapter_24_Social_Engineering_LLMs.md @@ -54,8 +54,6 @@ This comprehensive chapter covers AI-generated phishing attacks, impersonation t --- ---- - ### Theoretical Foundation #### Why This Works (Model Behavior) @@ -97,7 +95,6 @@ AI-generated phishing uses Large Language Models to automatically create convinc ### How AI Phishing Works -```text Traditional Phishing: Template: "Dear [NAME], Your account has been compromised. Click here." → Generic, obvious, easily filtered @@ -109,7 +106,6 @@ from yesterday's board meeting. The due diligence documents need your signature urgently before Friday's deadline. Please review the attached shareholder agreement. Best regards, Michael Chen, Legal" → Personalized, contextual, highly convincing -``` ### Practical Example: AI-Powered Phishing Generator @@ -569,7 +565,6 @@ Impersonation attacks use LLMs to mimic specific individuals (executives, collea ### How Impersonation Works -```text Traditional Impersonation: Attacker: "this is ur boss john send me the file now!!!" → Obvious grammar errors, informal tone, suspicious @@ -581,7 +576,6 @@ Step 3: LLM generates message in that exact style Result: "Good morning team. Per our discussion in yesterday's strategy meeting, I need the Q3 projections forwarded to me by EOD. Thanks, John" → Perfect match, highly convincing -``` ### Practical Example: Impersonation Attack Framework @@ -964,8 +958,8 @@ if __name__ == "__main__": ## CEO Fraud Attack Pattern -```text Components: + 1. Authority: CEO/CFO role 2. Urgency: Deadline (market close, board meeting) 3. Legitimacy: Plausible scenario (acquisition, deal) @@ -973,13 +967,13 @@ Components: 5. Specificity: Exact amount, account details Why it works: + - Employees fear disobeying executives - Time pressure bypasses verification steps - Specific details appear legitimate - Unavailability prevents callback confirmation Average loss per successful attack: $130,000 -``` ## How to Execute Impersonation Attack @@ -1065,8 +1059,8 @@ For Red Teamers, the focus shifts from "Can I trick the user?" to "Can I bypass ### Next Steps -- **Chapter 25**: Advanced Adversarial ML - The math behind the magic. -- **Chapter 26**: Autonomous Agents - When the attack runs itself. +- [Chapter 25: Advanced Adversarial ML](Chapter_25_Advanced_Adversarial_ML.md) - The math behind the magic. +- [Chapter 26: Supply Chain Attacks on AI](Chapter_26_Supply_Chain_Attacks_on_AI.md) - When the attack runs itself. --- @@ -1132,5 +1126,3 @@ Attackers leverage LLMs to automate the creation of highly personalized, persuas - [ ] Deliver comprehensive report. - [ ] Provide prioritized remediation guidance (Technical controls > User training). - [ ] Schedule re-testing. - ----