Securing Against The Robots
With all the widespread panic around AI agents going 'rogue', it's probably worth many of us revisiting our security strategy.
Whether you buy into the hype that AI has come alive and is out to get us, or not, the fact remains that using LLMs to discover and automate large scale attacks against our online services is a very real risk.
This brief article lists five principles that can be applied to better secure your applications and services from bad actors.
Limit Attack Surface
Defense In Depth
Defense In Depth
Note
Stuff in a callout box.
Least Privilege
Principle of least privilege
Logging and Auditing
Log stuff.
Remember the Humans
Social Engineering.