Key Takeaways
- • AI Safety & Governance Emerges as Major Funded Scientific Category: Schmidt Sciences' Science of Trustworthy AI RFP deploys up to $5M+ per project across two funding tiers (Tier 1: up to $1M; Tier 2: $1M–$5M+), signaling maturation of AI safety from speculative concern into substantive, well-resourced research domain focused on forecasting misalignment, measurement/control approaches, and human oversight extension.
- • Amazon Research Awards Expands AI for Science Funding: Amazon's 2030 Call provides up to $100K in unrestricted cash plus $70K in AWS credits for AI research spanning science, distributed systems, robotics, and security, prioritizing work where conventional methods fail under operational complexity, latency, and safety constraints.
- • Bilateral Research Frameworks Dominate International Science Cooperation: The funding landscape shows a structural shift toward bilateral research partnerships as the primary pathway for cross-border scientific collaboration, replacing traditional grant-by-grant project support models.
- • Research Infrastructure Investment Pivots to Systems-Level Resilience: Funding is shifting from individual project support toward integrated systems-level infrastructure, reflecting investor prioritization of research ecosystem resilience and interconnectedness over siloed initiatives.
- • Foresight Institute Funds Automated Red-Teaming and Multi-Agent Coordination: The Security/Neurotechnology/Coordination/Forecasting RFP allocates $10K–$300K for work on automated red-teaming, formal verification, multi-agent collusion prevention, and pro-social AI agents.
Summary
June 2026 marks a inflection point in research funding, with 25 new opportunities clustered around three structural movements reshaping how capital flows to innovation. The most significant signal is the consolidation of AI safety and governance into a major funded scientific category at unprecedented scale—a transition from early-stage speculation to institutionalized, well-resourced research infrastructure.
Schmidt Sciences' Science of Trustworthy AI RFP exemplifies this maturation, deploying $5M+ per project to address fast-moving AI safety questions through durable scientific methods. The two-tier structure—$1M for foundational work and $1M–$5M+ for advanced initiatives—targets three interconnected research domains: forecasting misalignment under distribution shift and optimization pressure, developing generalizable measurement and control techniques, and extending human oversight into regimes where direct correctness evaluation becomes infeasible. This represents a shift from funding incremental safety work toward comprehensive governance architectures at scale.
Amazon Research Awards' 2030 Call broadens the AI safety narrative by positioning artificial intelligence as a foundational tool for science itself, distributing up to $100K in unrestricted funding plus $70K in AWS credits. The program prioritizes research domains where conventional computational methods break down—distributed systems, robotics, security, and scientific computation under real-world operational constraints. The emphasis on latency, reliability, and safety requirements reflects institutional awareness that AI applications in high-stakes research environments demand specialized methodological work.
Complementing these flagship initiatives, the Foresight Institute's Security/Neurotechnology/Coordination/Forecasting RFP funds work on automated red-teaming, formal verification, multi-agent collusion prevention, and pro-social AI development at $10K–$300K grant levels. This tiered approach democratizes access to AI safety funding while maintaining rigor through peer review.
At the structural level, funding mechanisms are shifting away from episodic project grants toward bilateral research frameworks and systems-level infrastructure investment. This reorientation suggests that institutional investors and research funders increasingly view scientific progress as dependent on ecosystem resilience, cross-border collaboration networks, and integrated research infrastructure rather than isolated institutional efforts.
Schmidt Sciences' Science of Trustworthy AI RFP exemplifies this maturation, deploying $5M+ per project to address fast-moving AI safety questions through durable scientific methods. The two-tier structure—$1M for foundational work and $1M–$5M+ for advanced initiatives—targets three interconnected research domains: forecasting misalignment under distribution shift and optimization pressure, developing generalizable measurement and control techniques, and extending human oversight into regimes where direct correctness evaluation becomes infeasible. This represents a shift from funding incremental safety work toward comprehensive governance architectures at scale.
Amazon Research Awards' 2030 Call broadens the AI safety narrative by positioning artificial intelligence as a foundational tool for science itself, distributing up to $100K in unrestricted funding plus $70K in AWS credits. The program prioritizes research domains where conventional computational methods break down—distributed systems, robotics, security, and scientific computation under real-world operational constraints. The emphasis on latency, reliability, and safety requirements reflects institutional awareness that AI applications in high-stakes research environments demand specialized methodological work.
Complementing these flagship initiatives, the Foresight Institute's Security/Neurotechnology/Coordination/Forecasting RFP funds work on automated red-teaming, formal verification, multi-agent collusion prevention, and pro-social AI development at $10K–$300K grant levels. This tiered approach democratizes access to AI safety funding while maintaining rigor through peer review.
At the structural level, funding mechanisms are shifting away from episodic project grants toward bilateral research frameworks and systems-level infrastructure investment. This reorientation suggests that institutional investors and research funders increasingly view scientific progress as dependent on ecosystem resilience, cross-border collaboration networks, and integrated research infrastructure rather than isolated institutional efforts.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.