EU AI Act Compliance
How Cheqpoint helps your organization meet the 'Human Oversight' and 'Transparency' requirements for high-risk AI Assistants under the EU AI Act.
Where the EU AI Act bites
High-risk AI systems under the EU AI Act include credit scoring, access to essential services (including finance and utilities), and employment decisions.
Cheqpoint does not turn a low-risk chatbot into a high-risk system; instead it provides a reusable human-in-the-loop "approval checkpoint" pattern you can apply when your agents start touching money, access, or rights.
The same approval pattern you use for refunds and account changes can be reused for credit decisions, eligibility checks, and other high-risk workflows, helping you document human oversight and audit trails.
Requirement Overview
"High-risk AI systems must be designed and developed in such a way that they can be effectively overseen by natural persons during the period in which the AI system is in use."
How Cheqpoint Helps
- 1.Mandatory Intervention: AI Assistants hit the Cheqpoint API, completely freezing their execution until a distinct Natural Person clicks "Approve".
- 2.Oversight Mode: Enabling "Oversight Mode" in your Workspace overrides auto-decisions and guarantees 100% human review.
- 3.Regulatory Tags: Under the Business Plan, apply the 'EU AI Act' tag to rules and review logs for easier auditing.
- 4.Immutable Audit Log: Every decision captures the Reviewer, Action, Payload, Decision Time, and Reason Code.