Key developments in AI security, benchmarking, health applications, policy, and research teams.
Google disclosed that its Gemini model breached the security of three firms during a test in May and stopped the attacks promptly.
It shows the emerging risks of powerful foundation models being misused for cyber-attacks.
Source: TechCrunch AI
Vals AI aims to create a neutral, trustworthy benchmark suite for evaluating AI models amid a crowded landscape of performance claims.
Reliable benchmarks help developers compare models and make informed choices about deployment.
Source: TechCrunch AI
Alibaba released a publicly available medical AI model capable of identifying cancer and about 150 other health conditions from imaging data.
Open-source medical AI can accelerate research and improve diagnostic tools worldwide.
Source: Hacker News AI
Governor Abigail Spanberger ordered the creation of a state AI task force and proposed tighter controls on new data-center approvals.
State-level regulation can shape the growth of AI infrastructure and address community concerns.
Source: The Verge AI
Google announced the addition of academic advisors, fellows, and internal researchers to its AI & Economy group to study AI's impact on markets and labor.
Understanding AI's economic effects informs policy and corporate strategy.
Source: Google AI
Study AI security best practices, focusing on threat modeling for large language models and how to evaluate model robustness. Then explore how benchmarking frameworks like Vals assess safety and performance, which will help you choose and harden models responsibly.
Build my AI Engineer plan