UK, US strike landmark deal on AI safety testing’

Key Points:

  • UK and US partnering to co-establish testing procedures and guidelines on AI safety
  • Collaborative testing exercises and personnel exchanges planned
  • Positive industry reactions and focus on addressing AI safety challenges and governance frameworks


The UK and the US have forged a partnership following the EU’s recent approval of the risk-based AI Act. This collaboration aims to establish testing procedures and guidelines on AI safety by leveraging the capabilities and expertise of both countries. The plan involves developing tests for advanced AI models, creating tools for risk evaluation, and conducting collaborative testing exercises using publicly accessible models. Additionally, personnel exchanges between the UK and US AI Safety Institutes are on the agenda.

Michelle Donelan, the UK Secretary of State for Science, Innovation, and Technology, emphasized that ensuring the safe development of AI is a global issue that requires collective efforts. The collaboration between the UK and the US is seen as a way to address AI risks and unleash its vast potential to enhance lives.


Gina Raimondo, the US Secretary of Commerce, underscored that the partnership will accelerate work in managing risks related to national security and broader societal impacts. The announcement has been met with positivity in the European industry, with experts acknowledging the benefits for AI startups.


According to Ekaterina Almasque from the VC firm OpenOcean, the collaboration offers a significant advancement, particularly for AI startups dealing with safety and ethics complexities. She sees this partnership as a way to enable innovation, scalability, and competitiveness in the field.


Anita Schjøll Abildgaard, CEO of, views the partnership as crucial for establishing governance frameworks that keep pace with AI advancements. However, she warns that incorporating all stakeholders is essential to prevent fragmented approaches in AI safety standards and testing regimes. Abildgaard highlights Europe’s diverse ecosystem, spanning startups to industry leaders, as a valuable resource for informing international AI safety standards.


The UK-US partnership signifies a united effort to navigate AI’s challenges and potential, demonstrating a commitment to fostering innovation while prioritizing safety and ethical considerations.



Prompt Engineering Guides



©2024 The Horizon