Menu
In line with the Iraqi government's strategic directions in achieving digital sovereignty and enhancing technical independence, and under the supervision of the Supreme Committee for Artificial Intelligence, a specialized national team has been formed to undertake the task of:
Create a massive, multi-tasking, secure, and regionally competitive Iraqi LLM language model, built with Iraqi capabilities and national infrastructure.
This model aims to be the first large-scale, fully Iraqi LLM language model that serves the Arabic language in the local dialect and national content, and enhances the state's ability to control its data and smart systems.
1. Owning a sovereign AI infrastructure in Iraq.
2. Build a versatile language model that serves management, education, security, and health.
3. Reduce reliance on open or commercial foreign models.
4. Making AI tools available to Iraqi developers, universities, and research centers.
5. Promoting the privacy and digital sovereignty of government and public data.
1- A national multidisciplinary team was formed that includes:
* Experts in artificial intelligence and cloud computing.
* Specialists in Arabic language and local dialects.
* Engineers in Information Security and Natural Language Processing NLP.
* Representatives of Iraqi universities, the National Data Center, and the Media and Communications Authority.
2 - The team works under the direct supervision of:
* Advisor to the Prime Minister for Scientific Affairs and Artificial Intelligence.
* In technical cooperation with national and international research bodies.
Stage 1: Collect data and filter local content
* Collecting Iraqi textual data: media, education, legislation, dialects...
* Clean and standardize data within ethical and secure standards.
Phase 2: Infrastructure development
* Use national servers within the national data center.
* Equip scalable GPU clusters.
Stage 3: Building and training the model
* Design the model using open source libraries such as Hugging Face or DeepSpeed.
* Gradual training according to designated areas from 7 to 30B parameters.
Stage 4: Testing and Security
* Test the model in isolated environments to ensure it is free of biases or threats.
* Include safety layers and ethical filtering to protect use.
Phase 5: Launch and continuous improvement
* Launching a pilot version for ministries and universities.
* Release APIs for developers.
* Continuous development via user feedback.
* Iraqi custom language model IraqGPT or similar.
* A smart assistance system for government institutions in Arabic and Iraqi colloquial language.
* Legal and academic summarization and retrieval engine.
* Content analysis tools and corporate decision intelligence.
* Scalable protected language database.
1 . Support the project financially and technically within the 2025-2026 budget
2. Involving Iraqi universities in the stages of developing and training the model.
3. Enact national legislation regulating the establishment and operation of sovereign LLM models.
4. Establishing a national center for research in linguistic artificial intelligence.
A National Strategy for Artificial Intelligence seeks to propel a nation’s AI capabilities through investment in research, education, and economic growth, while ensuring ethical and responsible AI use.