[Revisited] Is it time to unite T&S and AI governance?

I'm Alice Hunsberger. Trust & Safety Insider is my weekly rundown on the topics, industry trends and workplace strategies that Trust & Safety professionals need to know about to do their job.

This week, I'm revisiting a piece I wrote in February 2025, which made the case for uniting Trust & Safety and AI governance. It merits circling back to because of everything that has happened over the last [pick your time horizon] related to AI governance. I'd love to know what you think.

On the topic of sharing your views, make sure you join EiM founder Ben Whitelaw and Georgia Iacovou from the Horrific/Terrific newsletter on Thursday 5 November for the latest Marked As Urgent event on the UK social media ban. They've got an amazing line-up of speakers; grab a spot while you can.

Here we go! — Alice

T&S and AI governance shouldn't sit in silos (updated for 2026)

I wrote the original version of this post almost 18 months ago as questions surfaced about how AI was changing the internet and why so few frontier labs were thinking about ethical deployment. So many things have shifted since then but the central point holds true so I've decided to revisit. Some of the references are updated but the core argument is the same. Enjoy — Alice.

Two major trends in the last couple of years have been AI's rapid expansion and the erosion of trust in content moderation.

More than ever, AI has become central to online platforms, powering everything from creative content to routine task automation, and, increasingly, AI agents that act on users' behalf. But it has also introduced new risks and harms. At the same time, we're continuing to see mass layoffs in Trust & Safety and widespread scepticism among company executives about content moderation and policy enforcement.

Together the message from the technology industry amounts to: we're going to invest in AI, but not in responding to its risks. If you ask me, that changes how we need to think and talk about Trust & Safety, as well as where it sits in company org charts.

Let me explain.

Getting behind "governance"

Trust & Safety teams were originally set up to handle human-centric issues and ensure community safety through guidelines and enforcement. This work continues today, but has expanded to include reviewing AI-generated content and using AI enforcement algorithms.

When I wrote the first version of this article, I titled it "T&S and AI ethics is not an ‘either or’ choice." At the time, the industry was thinking about where and how to incorporate AI into technology, and what the risks and ethical repercussions of those changes might be. AI ethics was still largely an academic and research discipline that was seeping into the consciousness of tech workers.

AI is now enmeshed in almost every product we use. Its risks are impossible to treat as hypothetical, and so platforms and enterprises have brought on governance people to manage them in practice.

These folks aren't academics; they're practitioners: pragmatic, operational, working with live products and real harms, much like T&S folks. And it's a fast-growing field. According to the IAPP, 77% of organisations are now doing AI governance work, and the privacy profession has started bringing AI governance, online safety, and cybersecurity together under the banner of "digital governance."

Of course, the more theoretical strand of this world hasn't gone away. The academic ethicists, and the frontier researchers worried about long-term or existential risk, are still at it. But that has become a different conversation from the practical, in-the-product work that T&S is part of.

Different means, same end goal

The other reason for uniting these fields is that T&S and AI governance are — at their core — both about putting company values into action. Yes, the systems, guardrails, and processes may differ, but both are used to build trust with users and mitigate risk. There are many examples of AI lab employees doing T&S-style work under a different name.

In practice, however, these teams and functions often work separately and in silos. This isn't good for users/consumers because it foregrounds the way that the company works over everything else. The IAPP puts this well in its 2025 digital governance report (emphasis mine):

"Interconnected risks are the new normal. Continuous innovation in the digital environment, whether it be the proliferation of Internet of Things devices or the rise of intuitive generative and agentic AI models, has shifted the way organizations perceive and operationalize their risk mitigation approaches. Today's risks span multiple domains and siloed approaches may prove to be fundamentally inadequate in addressing the new, interconnected and complex digital risk environment."

We've seen this disconnectedness play out over the last year, in both directions.