Transparent but Not Accountable
Why "open data" isn’t really open until people can actually use it and how we’re trying to fix that.
Governments, especially those with something to hide, often emphasize their transparency by publishing reports and data online. However, much of this information remains locked in inaccessible formats such as PDFs and scans, which neither people nor machines can easily process. This results in a performance of openness rather than genuine transparency, as information that cannot be meaningfully accessed, analyzed, or reused essentially remains hidden. True accountability requires structured, machine-readable data that enables public understanding and meaningful oversight, not just symbolic acts of disclosure. According to the OECD’s 2025 “Government at a Glance“ report, while regulatory frameworks in many countries promote transparency, the actual practice often falls short due to limitations in how accessible and usable the information is for the public. This gap challenges the effectiveness of transparency initiatives and highlights the need for improved data accessibility and usability to achieve real accountability (OECD, Government at a Glance 2025).
In reality, information that cannot be analyzed, compared, or searched cannot truly serve the public. A document uploaded to a ministry website might tick the "transparency” box, but if citizens, journalists, businesses, or researchers can’t extract meaning from it, it’s no different from being hidden. Accountability requires structure, and structure begins with data that can be read, interpreted, and reused.
At CDR, we’ve learned that being open isn’t enough. Real accountability starts when information becomes usable and machine-readable, when policy documents can be connected across institutions, and when citizens can understand the decisions made on their behalf. That’s why we’ve built tools like Referandom (we call it the Tinder for the Turkish Parliament) and its Instagram account to reach citizens while informing them about legislative processes in a sexier way, along with monitoring reports and datasets we share that help CSOs and researchers transform messy public records into structured data. These technologies, reports, and datasets allow them to ask better questions and challenge power with evidence rather than assumptions.
Policyful, our spin-out technology start-up, takes this work further by using AI to monitor parliaments, ministries, and regulatory bodies in real time. It brings together natural language processing, machine learning, and policy engineering to turn chaos into clarity by helping companies manage compliance more effectively, CSOs make data-driven advocacy easier, researchers analyze public policy trends and data over time. Because democracy isn’t just about access to information; it’s about understanding it, contextualizing it, and acting on it.