From 409532952b30c371c44441756bf8fda20241e0fc Mon Sep 17 00:00:00 2001 From: Rob van der Veer Date: Tue, 6 Feb 2024 17:40:32 +0100 Subject: [PATCH] Update ai_security_overview.md --- content/ai_exchange/content/docs/ai_security_overview.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/content/ai_exchange/content/docs/ai_security_overview.md b/content/ai_exchange/content/docs/ai_security_overview.md index e1265998..b5839d47 100644 --- a/content/ai_exchange/content/docs/ai_security_overview.md +++ b/content/ai_exchange/content/docs/ai_security_overview.md @@ -18,9 +18,9 @@ While AI offers powerful perfomance boosts, it also increases the attack surface - **Countermeasures in data science** through understanding of model attacks, e.g. data quality assurance, larger training sets, detecting common perturbation attacks, input filtering. ### Threat model -![AI Security Threats](/images/aisecthreat.png) +![AI Security Threats](/images/aisecthreats2.png) -![AI Security Threats and controls](/images/aisecthreatcontrols.png) +![AI Security Threats and controls](/images/aisecthreatscontrols2.png) ### Navigator diagram The navigator diagram below shows all threats, controls and how they relate, including risks and the types of controls.