How it works

Symbolic AI.
Deterministic output.

AiActs processes your system profile against EU AI Act provisions with a deterministic symbolic core. LLM-assisted interpretation is clearly separated from the primary report.

01

Questionnaire

An adaptive questionnaire with skip-logic. The number of questions depends on the breadth of your input and selected analysis mode. Generates a structured profile of your system for analysis.

02

Analysis engine

Your system profile is processed through rules covering all applicable legal norms. The engine fires rules based on your facts — fully deterministic, fully traceable.

03

Two compliance reports

Primary Report (Symbolic AI): Deterministic list of applicable norms — General Norms, Special Norms, Sanctions, Deadlines. Law data only, no interpretation. Every norm linked to the exact article. For legal teams and compliance officers.

Additional Report (LLM): Business-language interpretation for three audiences — executive summary with risk level, detailed legal analysis, and an action plan for the project manager. Available as HTML, PDF, Markdown.

04

Legal knowledge layer

A formalised EU AI Act knowledge layer with machine-readable rules, terms, and legal references. Traceability is being expanded across report formats.

aiacts_report_2026-03-28.html
System HR Screening Tool v2.3
Role Provider
Risk classification HIGH RISK
Annex III category 4(a) — Employment
Norms applied 47 applicable
Applicable obligations
Art. 9 — Risk management Art. 10 — Data governance Art. 13 — Transparency Art. 17 — QMS Annex IV — Technical docs
Logic trace · Rule R-0047
Art. 6(2) → Annex III(4)(a): System performs employment screening of natural persons → HIGH RISK classification triggered.
→ EUR-Lex: 32024R1689, Art. 6, §2
Reproducibility Symbolic core
LLM layer Separated from primary report