BeeKeeperAI®, Inc., a pioneer in privacy-enhancing multi-party collaboration software for AI, and cStructure, a leader in collaborative causal inference, announced a strategic collaboration to accelerate causal AI development. Their joint effort focuses on enabling scientists, health data stewards, and AI developers to build trustworthy AI models that uphold scientific rigor and healthcare compliance.
Causal AI: The Future of Trustworthy GenAI in Healthcare
This collaboration introduces a novel causal AI-centric approach that preserves patient data privacy while enabling transparency at scale. By modeling cause-and-effect relationships within large, sensitive health datasets, causal AI aims to create regulatory-grade, reliable insights. This approach enhances the trustworthiness and compliance of GenAI solutions, a key factor for adoption in life sciences.
Launching the Covid Causal Diagram DREAM Challenge
The first joint initiative, launching May 15, 2025, invites global scientists to analyze real-world NIH COVID-19 data in a privacy-enhancing environment. The challenge focuses on uncovering the causal impact of glucocorticoid treatment on 28-day survival rates for hospitalized patients.
Participants will develop Structural Causal Models (SCMs) using cStructure’s collaborative causal graph platform, assisted by integrated Large Language Models (LLMs) for domain expertise. Models will be evaluated securely in BeeKeeperAI’s EscrowAI environment, ensuring data privacy while simulating federated learning workflows.
Industry Leaders Speak
Dr. Michael Blum, CEO of BeeKeeperAI, highlighted the synergy:
“Our collaboration with cStructure leverages privacy-preserving tech and causal diagram tools to accelerate causal AI adoption for human health breakthroughs.”
Erick R. Scott, Founder of cStructure, emphasized:
“By combining secure collaboration and causal diagram development, we are bringing causal AI mainstream for science and healthcare.”
Gustavo Stolovitzky, Founder of DREAM Challenges, added:
“This effort blends privacy-preserving technology with global collaboration to push AI beyond prediction toward explanation, powering real medical breakthroughs.”
The Challenge Criteria
Models will be assessed on:
- Alignment with high-quality randomized controlled trial (RCT) results
- Proper confounder adjustment
- Plausibility of causal relationships
The BeeKeeperAI and cStructure collaboration marks a significant step toward making causal AI a cornerstone of trustworthy, privacy-compliant AI in healthcare and life sciences. The Covid Causal Diagram DREAM Challenge exemplifies how privacy-enhancing technologies and collaborative AI tools can drive transparency, regulatory confidence, and scientific progress.
