Giza Tech's Verifiable Recommendation System on RektHQ Reaches Two-Month Milestone

Giza Tech, in collaboration with investigative journalism platform RektHQ, has been successfully implementing a cryptographically verifiable recommendation system for the past two months, utilizing StarkWare's S-two prover. The initiative, highlighted by Raphael Doukhan, Lead zkML at Giza, aims to introduce unprecedented transparency into content delivery, allowing users to verify algorithmic integrity directly from their web browsers on RektHQ. This ongoing proof-of-concept addresses concerns about algorithmic bias on news sites and social platforms.

The core of this innovation lies in Giza Tech's application of zero-knowledge proofs (ZKPs), specifically STARKs, through its LuminAIR framework and StarkWare's S-two prover. This technology enables RektHQ to cryptographically prove that its content recommendations are generated precisely according to an open-source algorithm, without revealing underlying data. Users can independently verify these proofs client-side, ensuring the system operates as claimed and mitigating potential for hidden biases.

RektHQ, known for its in-depth analysis of decentralized finance (DeFi) and blockchain security, serves as an ideal partner for this pioneering effort due to its commitment to transparency and a security-conscious audience. The collaboration seeks to establish a new standard for trust and accountability in digital content, moving beyond opaque "black box" recommendation systems prevalent across major platforms. The project aims to empower users by shifting control from centralized entities.

Giza Tech, founded in 2022 and headquartered in Switzerland, specializes in bridging machine learning inference with blockchain technology through a trust-minimized framework. The company recently secured $3 million in funding to advance its mission of bringing verifiable machine learning to the blockchain, enabling decentralized and trustless AI applications. Their work with RektHQ exemplifies the practical application of verifiable AI in addressing critical issues like algorithmic transparency and content manipulation.

The long-term vision for this verifiable recommendation system extends beyond RektHQ, with plans to adapt the technology across various platforms, particularly social applications. As stated by Raphael Doukhan, "We believe that no news site or social platform should bias what you see." This initiative represents a significant step towards a more transparent and user-controlled internet, where the integrity of AI-driven systems can be mathematically proven.