By 
On Sunday, February 9, during AI Safety Connect 2025, an international event held in Paris alongside the Paris AI Action Summit,the Mohammed Bin Rashid School of Government (MBRSG), with the support of the Future of Life Institute (FLI), officially launched GRASP (Global Risk and AI Safety Preparedness).
RELATED: ITU’s AI for Good Global Summit 2025 to tackle the rise of autonomous AI
This platform is the first of its kind worldwide, mapping and categorizing AI safety tools and technologies designed to address the risks associated with artificial intelligence.
Cyrus Hodes, Fellow of Practice at MBRSG and Co-Lead of the SAFE Project at GPAI: “AI creates risks that companies, researchers, and policymakers must identify and address before they spiral out of control. GRASP provides a structured mapping of threats and solutions, helping key stakeholders adopt the right tools. The goal is to make technical, organizational, and regulatory solutions more accessible to mitigate risks associated with general-purpose AI and artificial general intelligence”.
This event brought together some of the world’s leading AI safety experts, including Yoshua Bengio, Stuart Russell, Max Tegmark, Xue Lan, Yi Zeng, and Jaan Tallinn. Together, they emphasized the need for a coordinated approach to AI governance, ensuring that technological advancements are developed and deployed with appropriate safeguards.
GRASP: A Groundbreaking Global Map of AI Risks and Solutions
Developed under the leadership of the Center for AI Security (CeSIA), in collaboration with Apart Research and the Beijing Institute for AI Safety and Governance, GRASP is the first comprehensive global mapping of AI safety solutions and technologies. It provides a structured approach to identifying major AI risks and the existing strategies to mitigate them.
GRASP’s work is part of the SAFE (Safety and Assurance of Generative AI) project, led by the Global Partnership on AI (GPAI) and now integrated into the OECD’s AI governance initiatives. GRASP and SAFE build upon the MIT AI Risk Repository, analyzing more than 80 types of AI risks and classifying over 200 solutions to address them.
Understanding and Anticipating AI Risks
As AI-related challenges grow, GRASP offers a structured framework to identify and address risks associated with advanced artificial intelligence systems.
Three Major Categories of Risks
- Malicious Use of AI: Automated cyberattacks, disinformation and deepfakes, autonomous weapons and bioterrorism, large-scale surveillance and censorship.
- Poor Design or Misalignment: AI developing unintended or deceptive behaviors, self-replicating systems and loss of control, the risk of superintelligent AI pursuing unintended goals.
- Unintended Societal Harms: Algorithmic bias and discrimination, impact on employment and economic inequality, influence on mental health and democracy, decline in human cognitive abilities.
A Key Resource for Policymakers and the AI Community
GRASP aims to be a reference tool for governments, businesses, and researchers, providing a comprehensive understanding of AI safety challenges. The platform enables users to explore and search for solutions tailored to specific risks and identify ongoing research initiatives and regulatory measures.
Key Features of the GRASP Platform:
- An interactive database categorizing risks and solutions.
- A modular approach linking each risk to concrete mitigation strategies.
- Detailed case studies illustrating both threats and countermeasures.
- An advanced navigation tool to filter information by risk type, application, or sector.
A Project Led by Recognized Experts
The GRASP project is led by Cyrus Hodes, Fellow of Practice at MBRSG in Dubai and Co-Lead of the SAFE project within GPAI. His role is to ensure coordination between stakeholders and to integrate GRASP and SAFE’s work within the OECD and GPAI.
Toward a More Responsible Global AI Governance
With the launch of GRASP, the international community now has a structured tool to address AI safety challenges. By facilitating access to existing solutions and promoting a proactive approach to risk management, GRASP aims to become a key reference for regulators, researchers, and industry leaders.
Explore the GRASP platform: https://aisafetypreparedness.com/explore
