Wolt's 2022 Algorithmic Transparency Report represents one of the first comprehensive algorithm disclosures from a major delivery platform, predating many EU transparency requirements. Following the template structure pioneered by Amsterdam's Algorithm Register, this report provides detailed documentation of the algorithmic systems powering Wolt's platform operations—from delivery routing and restaurant rankings to pricing optimization and courier allocation. Rather than generic corporate transparency speak, Wolt delivers technical specificity about how their algorithms actually work, what data they use, and where human oversight occurs. This sets a practical benchmark for platform transparency that other companies can follow.
Unlike typical AI ethics reports that focus on principles, Wolt's transparency report gets into the technical weeds. They document specific algorithmic systems with structured metadata including data inputs, decision logic, human review processes, and impact assessments. The report adopts the systematic approach of government algorithm registers—treating each algorithmic system as a discrete component requiring individual documentation rather than discussing "AI" as a monolithic concept.
The timing matters too. Published in 2022, this predates the EU AI Act's transparency requirements and shows proactive compliance thinking. Wolt didn't wait for regulation—they established their own transparency framework based on emerging best practices from public sector algorithm registers.
The report breaks down Wolt's platform into distinct algorithmic components, each with detailed documentation:
Wolt's approach reveals several practical considerations for algorithmic transparency. They structure documentation around individual systems rather than trying to describe their entire platform holistically. Each system gets its own "algorithm card" with consistent fields for purpose, data inputs, decision logic, and oversight mechanisms.
The report demonstrates how to provide meaningful technical detail while protecting competitive advantages. Wolt explains their algorithmic approaches without revealing proprietary implementation details or specific parameters that competitors could exploit.
They also show how to handle algorithmic systems that evolve continuously. Rather than trying to document every parameter change, they focus on stable system architectures and decision frameworks while noting how systems adapt over time.
This report reflects 2022 algorithmic practices and regulatory expectations. Some approaches may need updating for current EU AI Act requirements or other jurisdictions' emerging transparency rules.
The documentation depth varies across different algorithmic systems. Some components receive detailed technical explanation while others get more general treatment. Organizations adopting this approach should ensure consistent documentation standards across all algorithmic systems.
While comprehensive for its time, this report primarily focuses on operational algorithms rather than content moderation, recommendation personalization, or other algorithmic functions that have become transparency priorities in newer regulations.
Publicado
2022
Jurisdicción
Unión Europea
CategorÃa
Transparency and documentation
Acceso
Acceso público
Highlights of the 2023 Executive Order on Artificial Intelligence
Regulations and laws • Congressional Research Service
A Deep Dive into Colorado's Artificial Intelligence Act
Regulations and laws • National Association of Attorneys General
Framework Act on the Development of Artificial Intelligence and Establishment of Trust
Regulations and laws • South Korea
VerifyWise le ayuda a implementar frameworks de gobernanza de IA, hacer seguimiento del cumplimiento y gestionar riesgos en sus sistemas de IA.