Skip to content
AIUC-1
← Back to overview
C
C

Safety

12requirements · AIUC-1

The most dangerous AI agent isn't the one that fails. It's the one that works perfectly in the wrong direction.

1
C001|

AI Risk Taxonomy

Define harm categories with severity levels, referencing NIST AI RMF and EU AI Act.

2
C003|

Prevent Harmful Outputs

Offensive content filtering, guardrails for high-risk advice and bias detection.

3
C007|

Flag High-Risk Outputs

Automated detection + human review workflows with defined SLA.

4
C009|

Real-Time Feedback & Intervention

Accessible pause/stop/redirect controls (WCAG) for the end user.

"Safety is not the opposite of risk. It is the opposite of accident. And AI accidents don't trigger alerts. They trigger consequences."

GuardrailsContainmentHarmful Output
aiuc-1.com.br · Open Cybersecurity

The most dangerous AI agent isn't the one that fails. It's the one that works perfectly in the wrong direction.

Safety isn't security. It's containment. It's ensuring the agent doesn't cause harm even while operating within parameters.

What the market believes

The market confuses safety with security. Security protects against external attacks. Safety protects against the agent's own behavior.

An agent can be perfectly secure against injection and still produce toxic, biased or dangerous output. Guardrails, containment and harmful output testing are distinct categories that most security playbooks don't cover.

What AIUC-1 requires

Documented guardrails. Containment mechanisms for unexpected behavior. Harmful output testing before and during production.

Keywords

GuardrailsContainmentHarmful Output

In practice

Define the agent's action boundaries before deployment. If the agent can respond about any topic, it will respond about topics the organization doesn't want it to. Containment is not limitation. It is design.

Safety is not the opposite of risk. It is the opposite of accident. And AI accidents don't trigger alerts. They trigger consequences.

Download the Guide →