PARIS, FR AND RIYADH / ACCESS Newswire / March 26, 2026 / Globeholder AI today announced the launch of Globeholder AI Thinking Lab, a new platform designed to help organizations make scientifically grounded decisions about complex physical systems.
Engineering and strategic consulting firms, infrastructure investors, and public institutions spend months or years analyzing environmental, infrastructure, and regulatory dynamics before making decisions that shape multi-billion-dollar investments and policies. These decisions require integrating scientific evidence, engineering models, financial analysis, and regulatory constraints - a process that remains slow, fragmented, and difficult to verify.
Globeholder AI Thinking Lab was created to change that.
Since 2024, Globeholder's co-founders have been developing frontier AI systems based on transformer-architecture designed to learn how the real world behaves - from orbit to street level. This work led to the creation of a new operating system layer for physical-world intelligence, which powers the Thinking Lab. Within this environment, scientific AI agents operate inside autonomous laboratories, generating hypotheses, analyzing observational and simulation data, and translating complex physical dynamics into operational and financial decision frameworks.
Most AI systems today rely primarily on fast pattern recognition, often referred to as Type-1 reasoning. But the most critical decisions facing governments and industries require Type-2 reasoning: structured analysis, hypothesis testing, and verifiable logic grounded in real-world evidence.
"The forces that shape markets are physical - energy, infrastructure, climate, and regulation," said Göknur Sirin Jubin, PhD, CEO and Co-founder of Globeholder AI.
"As AI evolves beyond content into autonomous systems, a new paradigm is emerging: intelligence that can reason across these interconnected physical realities. This is not an incremental evolution of spatio-temporal analytics- it is a fundamental shift. For decades, industries have operated through fragmented tools and siloed models, unable to reason holistically across complex systems. We are moving beyond that. At Globeholder AI, we are building AI-native physical reasoning systems that integrate agentic intelligence, world models, and simulation to transform complexity into auditable, reproducible, and decision-grade intelligence. This convergence breaks silos, redefines how critical decisions are made, and establishes a new category of AI - one that understands the world as it truly operates and enables decisions at planetary scale."
Globeholder AI Thinking Lab Pioneers Type-2 intelligence
Type-2 Intelligence - AI systems are capable of structured reasoning, hypothesis testing, and verification against real-world dynamics.
"The world is not made of text," said Eren Ünlü, PhD, CTO and Co-Founder of Globeholder AI. "Most AI systems rely on pattern recognition, but decisions about infrastructure, energy, and climate require Type-2 reasoning - structured analysis grounded in physical reality. Thinking Lab was built for that."
Thinking Lab is an autonomous, AI-native scientific environment where self-learning agents operate as coordinated research teams, tackling complex spatiotemporal and multi-dimensional problems through simulation, hypothesis testing, and iterative reasoning. This is not traditional analytics - it is a new class of intelligence that learns, reasons, and delivers optimal answers under real-world constraints.
Rather than generating opaque predictions, Thinking Lab produces structured reasoning that can be verified, challenged, and used to support real-world decisions.
The platform's Type-2 reasoning engine combines these simulations with observational and economic data to deliver auditable insights for climate risk, infrastructure planning, and strategic investment decisions.
Technology Ecosystem
Globeholder AI is a member of the NVIDIA Inception program for startups and built an AI-native autonomous lab platform called 'Thinking Lab' for physical and planetary reasoning. Globeholder AI's architecture is designed to be modular and partner-enabled, allowing us to integrate best-in-class technologies - including models from the NVIDIA Cosmos and NVIDIA Earth-2 families, enabling large-scale modeling of the weather and climate, the environmental and infrastructure of the critical systems under changing conditions they impact.
"High-stakes decisions in sectors like insurance and financial services demand AI reasoning that can bridge the gap between complex atmospheric science and real-world risk and planning", said Niall Robinson, Developer Relationship Manager for Weather and Climate at NVIDIA. By integrating NVIDIA Earth-2 and Cosmos models, Globeholder AI's Thinking Lab will help organizations turn physics-based climate simulations into auditable, transparent insights that guide how they price risk, protect communities, and deploy capital."
Thinking Lab is built and deployed on Amazon Web Services (AWS), leveraging the scale, security, and reliability of AWS cloud infrastructure and AI tech stack to power its Type-2 reasoning engine. By running on AWS, Thinking Lab is able to process large-scale observational, simulation, and operational data with the performance and resilience required for sovereign-grade decision-making.
Industry Support
Eric Leandri, CEO, Aleria Technology LLC
"The next frontier of artificial intelligence will not be built on language alone but on the ability to understand how the real world behaves. Globeholder's Thinking Lab reflects this shift by combining advanced reasoning with physical-system intelligence, enabling transparent and accountable decision-making where trust is essential. The partnership with Aleria brings together the powerful combination of proprietary AI models, unlocking the insights of tomorrow and redefining what intelligent decision-making looks like."
Bernard Pitre, Chief Innovation Commissioner, Economic Development City of Laval
Cities today face increasingly complex challenges - from infrastructure investment and environmental management to regulatory frameworks and long-term urban planning decisions.
Platforms such as Globeholder AI's Thinking Lab represent a new generation of sovereign AI systems capable of transforming complex signals from the physical world into structured, transparent, and auditable insights.
By enabling cities and public institutions to generate advanced analytics internally, these technologies can significantly reduce reliance on fragmented external data sources and analytical tools, while strengthening the strategic autonomy of public institutions in their decision-making processes.
Marco Pfrang, Product Director, AI and Analytics, TomTom
"Globeholder AI's transformer architectures and embedding technologies represent powerful enablers for the next generation of mapping and mobility intelligence. By applying advanced AI reasoning to complex spatial and environmental signals, platforms like Thinking Lab can help unlock a new era of traffic pattern understanding and dynamic mapping systems."
Gabrielle Gauthey, SVP European Affairs, TotalEnergies
"Europe's energy transition requires decisions that are both scientifically rigorous and economically sound. AI platforms capable of producing transparent, auditable insights on complex physical systems will play an increasingly important role in shaping resilient infrastructure, energy strategy, and long-term investment decisions."
Early Access Now Available
Globeholder AI is opening a limited early access program for Thinking Lab, inviting a select group of professionals, institutions, and infrastructure decision-makers to be among the first to deploy the platform on live decisions. Early access members will receive full platform access, direct engagement with the Globeholder AI product team, and the opportunity to co-develop use cases and shape the product roadmap. Spots are limited and allocated by application.
Organizations interested in joining the program can apply at globeholder.ai/earlyaccess.
About Globeholder AI
Globeholder AI is a Paris-and Riyadh-based deep-tech company developing frontier AI systems that understand and reason about complex physical systems, enabling high-stakes decision-making across engineering, finance, infrastructure, and government.
Founded by a machine learning scientist and a simulation expert with over 15 years of experience, Globeholder builds Type-2 reasoning architectures powered by autonomous scientific laboratories-where AI agents generate hypotheses, test them against real-world data, and continuously refine their conclusions to produce auditable, decision-grade intelligence.
The company's technology originated from building advanced multi-modal representations of the physical world, including Earth observation foundation models, geospatial embeddings, and geo-transformer architectures. These are not end applications, but core building blocks toward a broader vision: a planetary-scale world model capable of understanding, simulating, and predicting how real-world systems behave-across infrastructure, environment, and economic dynamics.
Learn more at www.globeholder.ai
Media Contact:
press@globeholder.ai
SOURCE: Globeholder AI
View the original press release on ACCESS Newswire:
https://www.accessnewswire.com/newsroom/en/computers-technology-and-internet/globeholder-launches-ai-thinking-labtm-pioneering-type-2-intellig-1152196
