CBE guides Israeli AI companies to compliance with the EU AI Act, ISO 42001 and ISO 42005 — from AI risk assessment to full certification.
Why CBE for AI?
The EU AI Act classifies every AI system by its risk level. CBE sorts your AI and builds a tailored compliance plan.
UNACCEPTABLE RISK
AI systems that are completely banned in the EU.
Examples: Social Credit Scoring, biometric emotion recognition, exploitation of vulnerabilities
HIGH RISK
Requires Technical Documentation, Risk Management and Conformity Assessment.
Examples: AI for credit, hiring, medical diagnosis, law enforcement, education
LIMITED RISK
Mandates transparency — disclosing that the user is interacting with AI.
Examples: Chatbots, Deep Fakes, Emotion Recognition
MINIMAL RISK
No mandatory requirements — but adopting a code of ethics is recommended.
Examples: Spam filters, chess AI, content recommendations
GPAI — General Purpose AI Models
Models like GPT, Claude, Gemini — subject to transparency and copyright requirements. Models with Systemic Risk (over 10²⁵ FLOPS) are subject to stricter requirements.
CBE's AI Compliance Assessment Tool
Automated assessment for EU AI Act and ISO 42001 — instant risk-level identification and compliance plan.
CBE specializes in tailored AI Governance guidance for every sector.
AI algorithms for credit and insurance are classified as High Risk. CBE guides compliance and ISO 42001 + DORA.
AI for diagnosis and treatment — High Risk. ISO 42001 + ISO 13485 + HIPAA. CBE builds a medical AI Governance framework.
AI systems for CV screening and candidate ranking — High Risk. Strict anti-bias requirements and GDPR for personal data.
Chatbots and GenAI — Limited Risk. Transparency duty to the user + GPAI compliance.
AI for predictive maintenance and quality control — High Risk for critical infrastructure. ISO 42001 + ISO 27001 + NIS2.
ADAS and autonomous driving — High Risk + TISAX + UNECE requirements.
AI in policing and public services — High/Unacceptable Risk with stringent requirements.
BI tools with AI — Minimal Risk. ISO 42001 and GDPR for personal data.
August 2024 — official publication
EU AI Act published in the Official Journal and entered into force.
February 2025 — prohibitions
Unacceptable Risk AI is banned from operating in the EU.
August 2025 — GPAI
GPAI requirements in effect — transparency, copyright, systemic risk.
August 2026 — High Risk AI
Every High Risk AI requires Conformity Assessment and registration in the EU Database.
2027 — safety products
AI in products covered by existing safety regulations.
35M€ / 7%
Unacceptable Risk violations
Using AI that is completely banned.
15M€ / 3%
High Risk & GPAI violations
Marketing High Risk AI without Conformity Assessment.
7.5M€ / 1.5%
Incorrect info to authorities
Providing incorrect information to the European AI Authority.
CBE guides all the standards and regulations relevant to AI — under one roof.
Yes. The AI Act is extra-territorial — it applies to any organization that develops, imports or uses AI offered in the European market. An Israeli company marketing AI to the EU must comply with the Act.
Not sure what applies to your AI? CBE performs an AI Risk Classification — within 48 hours you'll know exactly what the requirements are.
30 years of experience guiding ISO certification and regulation. We built Israel's first AI Compliance tool and guide AI companies from day one to certification.
Prefer to talk?
03-5105232Address
Nationwide service