and the distribution of digital products.
Can Blockchain and AI Finally Trust Each Other? Flare and Google Cloud’s Hackathon Says Yes
How can blockchain handle complex AI computations without losing its core promise of transparency? The Flare x Google Cloud Hackathon, held from March 7-9, 2025, at UC Berkeley’s California Memorial Stadium, tackled this question head-on. Over 460 developers and researchers from institutions like UC Berkeley, the University of Waterloo, and ETH Zurich participated, showcasing how blockchain and AI can integrate using verifiable off-chain computation. The Flare x Google Cloud Hackathon, held from March 7-9 in collaboration with Blockchain at Berkeley, demonstrated how verifiable AI can be seamlessly integrated with blockchain technology.
\
Blockchain excels at decentralized consensus but falters with the heavy data processing AI demands. The event introduced a solution: Google Cloud’s Confidential Space, built on Trusted Execution Environments (TEEs). TEEs process complex tasks off-chain in secure hardware enclaves, producing cryptographic attestations verified on Flare’s blockchain. This ensures computations remain tamper-proof. Flare’s built-in protocols, the Flare Time Series Oracle (FTSO) and Flare Data Connector (FDC), provided data integrity and provenance, critical for AI applications.
\ Further, one of the key challenges in blockchain technology is achieving rich data computation while maintaining transparency and security. Traditional blockchains excel at decentralized consensus and deterministic execution but struggle with large-scale parallel processing, particularly for AI-intensive workloads. This hackathon sought to overcome that limitation by leveraging Google Cloud Confidential Space, a secure enclave built on Trusted Execution Environments (TEEs).
\ Confidential Space ensures hardware-enforced confidentiality and generates cryptographic attestations, which can be verified on the Flare blockchain. This process guarantees that computations occur in a tamper-proof environment, preserving both data integrity and security. Flare’s enshrined data protocols, such as the Flare Time Series Oracle (FTSO) and the Flare Data Connector (FDC), played a crucial role in ensuring data provenance and integrity verification for AI applications.
\
With over 460 participants competing across four tracks, the hackathon produced verifiable AI solutions ranging from AI-powered decentralized finance (DeFi) applications (DeFAI) to consensus learning models. Every project utilized Confidential Space instances running on AMD’s Secure Encrypted Virtualization (SEV) platform with virtual Trusted Platform Module (vTPM) attestations. These attestations provided cryptographic proof that all AI computations were executed within an isolated, trusted environment.
\ Rene Kolga, Product Manager at Google Cloud, expressed enthusiasm about the developers' engagement with the Confidential Space secure enclaves, stating:
\ “We're thrilled to see developers using our Confidential Space secure enclaves to build trusted AI solutions. Projects like 2DeFi and the Flare AI Kit’s launch are exciting steps towards a future where AI and blockchain ensure data integrity and computational assurance.”
2DeFi: Bridging Traditional Finance and DeFi with AIOne of the outstanding projects from the hackathon was 2DeFi, an AI-powered solution developed by University of Waterloo students Alex and Hitarth. The project won first place in the AI x DeFi (DeFAI) track and secured second place in the RAG Knowledge track.
\ 2DeFi aimed to simplify the transition from traditional finance (TradFi) to decentralized finance (DeFi) by integrating Google Gemini AI. Users could upload screenshots of their Robinhood portfolios, which Gemini 2.0 then analyzed to assess risk tolerance. The AI-generated risk profile was mapped to an automated DeFi strategy, incorporating staking and liquidity provision on Flare.
\ To streamline onboarding, 2DeFi also introduced embedded wallets, allowing users to create Google login-based wallets on Flare, removing the complexities of traditional Web3 user experiences.
Flare AI Kit: Empowering Verifiable AI on BlockchainThe hackathon also marked the official kickoff of Flare AI Kit, an open-source software development kit (SDK) designed for building verifiable AI agents on Flare. The kit leverages Google Cloud Confidential Space to provide developers with secure, attestable AI execution environments, ensuring AI safety, provenance, and verifiability.
\ Hackathon participants demonstrated the potential of Flare AI Kit as a foundation framework for AI-powered blockchain applications. Moving forward, Flare AI Kit is expected to evolve into a production-ready framework, further enhancing functionality, usability, and integration options. The ultimate goal is to make Flare AI Kit available on the Google Cloud Marketplace, where enterprises and developers can access it as a turnkey solution.
The Future of Verifiable AI on BlockchainThe success of the Flare x Google Cloud Hackathon is just the beginning. As developers and researchers continue to push the boundaries of AI and blockchain convergence, projects like Flare AI Kit will play a crucial role in shaping the future of verifiable AI applications.
By integrating Confidential Space TEEs, AI, and the Flare blockchain, Flare AI Kit is set to redefine how secure, verifiable AI applications are built, deployed, and trusted globally. Developers looking to get started can explore the Flare Developer Hub for in-depth documentation and resources.
\ Don’t forget to like and share the story!
\