Skip to main content
AI & Technology

EU AI Act — August 2026 Deadline for High-Risk AI: Who Needs to Comply?

EU AI Act กำลังจะบังคับใช้ข้อกำหนด High-Risk AI เต็มรูปแบบในสิงหาคม 2026 ด้วย extraterritorial scope ที่ครอบคลุมทั่วโลก โทษปรับสูงสุด 7% ของรายได้รวม — องค์กรไทยที่มีลูกค้า EU ต้องเตรียมตัว

5 Mar 20266 minLegalNodes
EU AI ActAI GovernanceHigh-Risk AICompliance

Countdown: August 2026

The EU AI Act, the world’s first comprehensive AI law, is approaching its most important enforcement phase: from 2 August 2026, the requirements for High-Risk AI Systems will become fully applicable.

Crucially, this law has extraterritorial scope — meaning that regardless of where your company is located, if your AI “affects” individuals in the EU, you must comply.

What Is High-Risk AI?

The EU AI Act classifies AI into 4 levels:

  1. Unacceptable Risk — strictly prohibited (e.g. social scoring, real-time biometric identification in public spaces) — already in force since February 2025
  2. High Risk — subject to strict compliance requirements — 2 August 2026
  3. Limited Risk — subject to transparency obligations
  4. Minimal Risk — no additional regulatory requirements

Examples of AI That Qualify as High-Risk

Category Examples
Employment AI resume screening, AI interviews, AI performance evaluation
Finance AI loan approval, AI credit scoring
Education AI exam grading, AI admissions decisions
Healthcare AI disease diagnosis, AI treatment recommendations
Legal AI-assisted judicial decision-making, AI crime risk assessment
Infrastructure AI controlling electricity, water supply, transportation systems

Compliance Requirements

Organizations that develop or use High-Risk AI must:

For Providers (developers/distributors)

  • Establish a Risk Management System throughout the AI lifecycle
  • Implement Data Governance to ensure the quality of training data
  • Prepare Technical Documentation explaining how the AI system works
  • Maintain Logging systems for traceability and auditability
  • Provide sufficient information to the Deployer (user organization)
  • Design systems to enable Human Oversight
  • Conduct a Conformity Assessment before placing the system on the market
  • Register in the EU Database for high-risk AI

For Deployers (user organizations)

  • Use AI in accordance with the provider’s instructions
  • Ensure human oversight in operation
  • Verify the input data fed into the AI system
  • Retain logs as required
  • Conduct a Fundamental Rights Impact Assessment (FRIA) in certain cases

Penalties

The fines under the EU AI Act are substantial:

  • Prohibited AI → fines of up to EUR 35 million or 7% of total worldwide turnover (whichever is higher)
  • Non-compliance with High-Risk AI requirements → fines of up to EUR 15 million or 3% of total turnover
  • Providing false information → fines of up to EUR 7.5 million or 1% of total turnover

Why Should Thai Organizations Pay Attention?

Although the EU AI Act is European legislation, its extraterritorial scope creates cross-border impact:

  • Thai exporters — if your products or services use AI and are sold in the EU, you may fall within scope
  • Thai companies with EU customers — AI used to serve customers in Europe (e.g. SaaS, platforms) must comply
  • A model for Thai legislation — Thailand’s draft AI law draws on many concepts from the EU AI Act
  • The Brussels Effect — just as GDPR became a global benchmark, the EU AI Act is likely to drive higher standards worldwide

Key Timeline to Remember

Date What Happens
Feb 2025 Prohibitions on unacceptable-risk AI take effect
Aug 2025 General-Purpose AI (GPAI) requirements take effect
Aug 2026 High-Risk AI requirements take effect
Aug 2027 Additional requirements for High-Risk AI under Annex I

What You Should Start Doing Today

Whether or not your organization is directly affected by the EU AI Act, preparing now will be beneficial:

  1. AI Inventory — identify what AI your organization uses and classify risk levels under the EU framework
  2. Gap Assessment — assess what your current AI systems are missing relative to the requirements
  3. Documentation — begin preparing technical documentation for critical AI systems
  4. Human Oversight — verify that there is human-in-the-loop control for AI making significant decisions
  5. Data Governance — review the quality of data used to train and feed AI systems

AI regulation will become the norm, not the exception — organizations that prepare today will not have to scramble tomorrow.


Source: EU AI Act 2026 Updates: Compliance Requirements and Business Risks — LegalNodes, Implementation Timeline — EU AI Act

"Empowering Innovation,
Transforming Futures."

ติดต่อเราเพื่อทำให้โปรเจกต์ของคุณเป็นจริง