Beyond disclosure: transparency in an age of AI

Democracy rests on visibility: citizens must see their interests treated equally and have access to the knowledge needed for informed debate. Freedom of Information (FOI) laws in the late 20th century advanced this ideal mostly through reactive transparency — requiring citizens to request information. Later, open data initiatives pushed for proactive transparency.

But transparency has never been a panacea. Studies show FOI increased openness and accountability but did not automatically build trust. If governance is weak, more visibility can actually reduce trust. Transparency reveals quality but it cannot replace it.

Today, AI raises the stakes. Decisions once made by humans are now shaped by algorithms, often behind corporate firewalls. Companies publish AI ethics statements, but stakeholders still ask: “Can we trust how you use data and algorithms?” The same principle that justified opening government files now demands corporate clarity on AI systems and data practices. The arena has shifted but the principle remains the same.

FOI is mostly about access to documents. AI demands access to logic: how automated systems make decisions that affect people’s lives. Algorithms now influence corporate strategy, hiring, and performance evaluations.

Meanwhile, transparency is no longer only top-down. As columnist Sarah O’Connor notes, we live in a world of “Little Brothers” — employees and peers monitoring each other through digital tools.

This creates a paradox: the more data we expose, the less trust we may generate. For employees, constant visibility can feel like surveillance. For customers and stakeholders, dashboards and disclosures often generate noise, not clarity. Transparency without explanation risks becoming a source of anxiety rather than accountability.

To move beyond disclosure, companies must provide clarity, context, and conversation. That means being specific, explainable, and auditable. Four practices matter most:

• Evidence: What systems you use, what data you collect, and why.
• Explanation: How decisions are made and what trade-offs exist.
• Examination: Independent checks and meaningful metrics.
• Engagement: Two-way dialogue with employees, customers, and society.