How Trust & Safety teams can do more with less

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

• The US Supreme Court case worth paying attention to
• How Trust & Safety leaders at smaller companies are changing strategy as a response to being asked to do more with less.
• How to improve your written communication.

Get in touch if you'd like your questions answered or just want to share your feedback. Here we go! — Alice

A landmark US legal case

Oral arguments for Murthy v Missouri are happening today in the US Supreme Court. For those who don't know much about it, the case centres around the concept of "jawboning," in which government officials pressure or coerce social media platforms to take down user-generated content. It's important in the context of the US First Amendment but could have broader implications about how democratic governments address mis and disinformation without infringing citizen's speech.

Here are some links to get you up to speed:

• What's at stake in Murthy v Missouri? (Tech Policy Press)
  Why? Tech Policy Press sent out this helpful email full of links this morning ahead of the Murthy v Missouri oral arguments in the US Supreme Court today. If you only have time to read one of their articles, it should be this background piece.
• An FAQ on Murthy v. Missouri Ahead of Oral Arguments Tomorrow in the Supreme Court (Klonikles, Substack)
  Why? This is another comprehensive article that goes over the basics of the case. It also contains even more links at the bottom to get you up to speed.
• Supreme Court Case Could Be Disastrous for Detecting Election Misinformation (Brennan Center)
  Why? This piece explains how Murthy v Missouri is part of a larger political effort to silence those working on election misinformation.
• Jawboned (Knight First Amendment Institute)
  Why? While not new, this essay gives a comprehensive look at jawboning from the perspective of two former Facebook employees. It discusses the pros and cons and outcomes of jawboning, and suggests a path forward.

How Trust & Safety teams are changing

One of the things I've been enjoying about my new job is that I get to talk to people at a variety of companies about Trust & Safety. Based on conversations I've had with leaders, as well as resources I link to, here's what I think is changing in Trust & Safety at small and mid-sized companies:

AI is expanding the definition of "Trust & Safety" work

For many leaders in small to mid-sized companies, T&S work has primarily been centred around policy and operations to manage user-generated content: create a set of rules, manage a team of moderators, remove bad content (or bad actors). T&S leaders at even the smallest companies are now becoming more systems-focused and cross-functional. Policy and operations are a small piece of a large set of responsibilities.

Much of the work I've done over the last few years has been implementing machine learning models and thinking about AI ethics, and that's true of many T&S leaders. T&S teams are responsible for detecting and removing LLM-generated illicit content, but also are red teaming, testing, evaluating, fine-tuning, and overseeing AI models from start to finish. As product teams design AI into almost everything, T&S teams are tasked with keeping users safe from the platform's own products as well as from other users.

Read more: Reflections on a Month in Silicon Valley (Anchor Change)

Safety regulation is just beginning

T&S leaders hope that regulation incentivises investment in work that will make a difference for their teams, but many companies are taking a "wait and see" approach, or rolling out MVP versions of projects to be compliant. Dialogue is happening between policy makers and T&S experts, but we need to have more opportunities for regulators to talk to T&S leaders so they can align on what changes will actually make a difference.

Read more: 1 year at the Integrity Institute: what I've learned (Abby Lawson, Integrity Institute)

There's also more harmful content out there

We're seeing rising levels of child sexual abuse material and non-consensual intimate imagery. An unprecedented number of people around the world will vote in elections this year while disinformation is easier than ever to generate. Extremism is on the rise. There's more to do than ever.

Read more: The State of Trust & Safety 2024 (ActiveFence)

Leaders are bringing in vendors sooner and for smaller projects

Historically, leaders have invested in T&S tech and outsourced labour only when their userbase reaches scale. However, layoffs across the industry have resulted in pressure to keep team lean, even at the smallest companies. Small T&S teams can't do it all, and so instead of expanding internal teams, leaders are looking to vendors sooner. Easy tasks are automated through SaaS tech and AI, more complicated tasks are assigned to humans, and the task of organising and optimising the strategy is left in-house.

One T&S tool to rule them all

Safety tech as an industry is becoming more mature and reliable — meaning its less risky for practitioners to take a chance on integrating tools into mission-critical services. Eventually, I expect there will be a well-developed Zendesk or Salesforce equivalent that dominates the market and is the go-to SaaS for moderation.

Because of regulatory pressure for moderation in many languages and the need for quick response times 24/7 for illegal content, even small companies are looking to use a business process outsourcing (BPO) company for some portion of moderation needs (Full disclaimer: I recently started a new job for PartnerHero, which does T&S outsourcing, but have worked with different BPOs in my career).

Instead of bringing on a BPO only when hundreds (or thousands) of moderators are needed at scale, I'm seeing startups considering partnering with a BPO right from the beginning. Team sizes may be smaller as the easy work gets automated, but the expertise will deepen.

Read more: Trust & Safety Market Research Report (Duco Experts)

I predict that due to these changes, internal T&S teams — or, in some cases, what remains of them — will become more fully integrated into their companies. As T&S leaders spend more time on "proactive" work , such as product and AI projects, I hope that T&S will no longer be seen as a "cost centre", but instead as a strategic partner to many departments. We'll see T&S teams taking more responsibility, including Chief Safety Officers who report directly into CEOs. Internal T&S teams may not expand a great deal, but the impact they have will.

Job hunt

For any Trust & Safety professional, an important skill is being able to clearly and succinctly explain complicated scenarios. Knowing how to write a good memo, brief, or request is critical. Making the rounds in my internet circles this week was Make better documents, Anil Dash's guide to, well, making better documents. It's well worth a read, whether you're writing for the C-suite, creating LinkedIn content, or sprucing up your cover letter.

It reminded me of these excellent pieces: writing tips from the president's daily brief and Q&A on writing for execs that came out of a TrustCon workshop last year.

Finally, if you've been laid off but would like to go to TrustCon, the TSPA is offering scholarships and 50% discounted tickets. Applications are open now. You can also apply for a free extension on your TSPA membership.

Also worth reading

How to make Chinese platforms American (Dagens Tech podcast)
Why? The US House voted to ban (or force a sale of) TikTok. What I see left out of commentary on this is that there is a precedent: Grindr's Chinese parent company was forced to sell to US investors in 2020 due to national security concerns. This podcast interviews CEO Jeff Bonforte who was part of the investing group that bought Grindr and "made it American" (Jeff brought me on to lead T&S at Grindr just a couple of months after the sale).

Can Reddit survive its own IPO? (Wired)
Why? An in-depth read about the importance of community for Reddit, and it's colourful history with volunteer moderators.

Coping in the queue: tips for managing AI anxiety in customer support (Help Scout)
Why? A great read for anyone who is worried that AI might take their job, especially for those in frontline positions like customer support or content moderation. This article has some great tips about how to think about the future of frontline work and your own career.