The GNI Has a Clear Message: Human Rights Are Losing Ground to “Safety”

Key Takeaways

• The Global Network Initiative (GNI) and the Digital Trust & Safety Partnership (DTSP) have published a summary report from their April 2026 Rights & Risks Forum in Dublin, attended by more than 90 experts from around the world.
• Participants warned that contested definitions of "systemic risk" in regulatory discourse are eclipsing human rights concepts, and that framing digital spaces as inherently dangerous may be providing political cover for government overreach.
• The Forum's more than 50 recommendations include a call for governments to explicitly center democratic values and international human rights standards in the design and enforcement of online safety regulations.
• Experts also warned that the proliferation of online safety laws with distinct risk management requirements is leading to the restriction or removal of lawful content, not just harmful material.

The Global Network Initiative and the Digital Trust & Safety Partnership have released a summary report from their Rights & Risks Forum, held in Dublin in April 2026, according to a post published on the GNI's website on July 1. More than 90 experts attended, including company practitioners, civil society representatives, academics, independent researchers, and international organization staff. Government officials and regulators were deliberately not invited, in order to encourage candid discussion.

The Forum's central focus was on how technology companies can assess and mitigate risks to fundamental rights, specifically freedom of expression and privacy, within a regulatory environment that is growing more complex and more aggressive by the year. The headline finding is one I think deserves wider attention than it will probably get: in the current regulatory discourse, contested definitions of "systemic risk" are threatening to eclipse human rights concepts entirely.

Put plainly, governments around the world are passing online safety laws that frame the internet as a source of danger, and that framing is increasingly being used to justify restrictions that go well beyond what the stated safety goals require. The Forum's participants warned explicitly that "the perception of digital spaces as inherently risky may provide cover for political abuse and government overreach."

The Gap Between Goals and Outcomes

One of the most important observations in the report is about implementation. The Forum found that insufficient attention is currently being paid to how risk-based regulatory approaches actually work in practice. 

The proliferation of online safety laws, each with its own distinct risk management requirements, each pulling platforms in slightly different directions, isn’t producing a safer internet. It’s producing an internet where lawful content is increasingly restricted or removed because platforms, faced with legal uncertainty and compliance costs, default to over-moderation.

This is a dynamic I think most people don't see clearly because the outcomes look superficially like safety. A post gets removed. An account gets suspended. A platform restricts access in a particular country. From the outside, these look like content moderation decisions. What the Forum is pointing out is that many of them are actually the downstream effects of regulatory frameworks that are poorly designed, inconsistently applied, and structurally incentivize removal over retention of anything that could be interpreted as risky.

The report also highlighted how shrinking civic space (the global trend of civil society organizations, journalists, and activists facing increasing pressure) compounds this problem. When the people best positioned to identify over-moderation and rights violations are themselves under pressure, the feedback loops that should correct regulatory overreach stop functioning.

What the Experts Recommend

The Forum produced more than 50 recommendations. Three stand out to me as particularly important. First: governments should explicitly center democratic values and international human rights standards within the text, design, and enforcement of online safety regulations, not just invoke them as aspirational framing while building frameworks that functionally undermine them. 

Second: companies should ground their risk management approaches in international human rights law and the UN Guiding Principles on Business and Human Rights, independently of their specific legal compliance obligations. 

Third: experts in business and human rights need to do more to clarify how risk management frameworks can be grounded in human rights, because right now, the language of risk is winning and the language of rights is losing.

I find this report significant not because it breaks new ground, but because of who is saying it and in what context. This isn’t a fringe advocacy position. It’s the conclusion of over 90 practitioners, researchers, and civil society experts who work inside these systems every day. Their collective warning that the regulatory environment is drifting away from rights protection and toward control is one that people who care about a free internet should take seriously. The direction of travel matters, even when individual laws seem reasonable in isolation. The destination those laws are collectively building toward is the question worth asking.