From the course: CompTIA SecAI+ (CY0-001) Cert Prep
Unlock this course with a free trial
Join today to access over 25,500 courses taught by industry experts.
Network and API access controls
From the course: CompTIA SecAI+ (CY0-001) Cert Prep
Network and API access controls
Controlling access to an AI system's network and APIs is one of the most important steps in keeping it secure. These endpoints act as a front door to the model, allowing users, systems, or applications to send requests and receive responses. If the door is left unguarded, attackers can exploit this to steal data, overload the system, or manipulate the model's behavior. Strong endpoint and endpoint and network access controls ensure that only authorized requests reach the model and that all communication remains private and tamper-proof. In a previous video, we covered the importance of authentication, authorization, and usage limits as part of model access control. Those same principles apply here, but at a network and API level, where they serve as the first set of checkpoints for all incoming traffic. Each request must be verified to confirm who is making it and what they are authorized to do. Public facing APIs typically require tokens or API keys, while internal systems rely on…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.
Contents
-
-
(Locked)
The AI lifecycle1m 39s
-
(Locked)
Business alignment in the AI lifecycle1m 43s
-
(Locked)
Data collection2m 20s
-
(Locked)
Data preparation3m 15s
-
(Locked)
Model development and selection2m 13s
-
(Locked)
Model evaluation and validation2m 29s
-
(Locked)
Model deployment and integration3m 25s
-
(Locked)
Monitoring and maintenance3m 19s
-
(Locked)
-
-
(Locked)
Manipulating application integrations4m 8s
-
(Locked)
AI supply chain attacks2m 4s
-
(Locked)
Insecure plug-in design2m 9s
-
(Locked)
Insecure output handling1m 23s
-
(Locked)
Output integrity attacks2m 8s
-
(Locked)
Model denial of service1m 31s
-
(Locked)
Excessive agency1m 33s
-
(Locked)
Overreliance1m 34s
-
(Locked)
AI hallucinations1m 4s
-
(Locked)
-
-
(Locked)
Monitoring prompts and responses2m 51s
-
(Locked)
Log monitoring4m 30s
-
(Locked)
Rate and cost monitoring5m 1s
-
(Locked)
Auditing for AI hallucinations3m 33s
-
(Locked)
Auditing for accuracy3m 29s
-
(Locked)
Auditing for bias and fairness4m 35s
-
(Locked)
Auditing access and security compliance3m 48s
-
(Locked)
-
-
(Locked)
Responsible AI5m 29s
-
(Locked)
AI risks2m 23s
-
(Locked)
Introduction of bias2m 37s
-
(Locked)
Accidental data leakage2m 53s
-
(Locked)
Reputational loss2m 11s
-
(Locked)
Accuracy and performance of the model2m 22s
-
(Locked)
Intellectual property risks3m 31s
-
(Locked)
Autonomous systems2m 27s
-
(Locked)
Shadow IT and shadow AI1m 48s
-
(Locked)
Awareness training2m 21s
-
(Locked)