My must-read election integrity resources

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:

• How to navigate the many guides to election integrity
• Why AI tools make me less productive, not more

Get in touch if you'd like your questions answered or just want to share your feedback.

And if you want to say hi in person, I'll be at Marketplace Risk next week, speaking on a couple of panels and soaking up all the news and insights. Maybe I'll see you there?

— Alice

Bookmark these election integrity guides

Maintaining information integrity around global elections is a difficult balance with high stakes. Under-enforcement and over-enforcement are both concerns, and so content moderation decisions must be weighed against any potential real-world harm that may happen. It's easier said than done, especially when there are elections happening globally.

However, there is help out there. Over the last year, we've seen a host of new additions to the already robust list of resources and advice available to T&S practitioners. All of these guides can get overwhelming, but I'm here to help! Here are my favourites, organised into a series of reminders/tips:

Determine your risk

Different organisations platforms have different appetites for risk during elections. So starting with a series of questions to help you figure that out what the looks like is time well spent.

• The Integrity Institute guide released last year suggests red teaming to determine risk levels, linking to this doc with some great questions to ask in red team exercises, while the Elections Playbook for Startups suggests tabletop exercises.
• This Election Cycle Tracker by Anchor Change can help you see what elections are coming up, which will help you prioritise.

Define what good looks like

Like any T&S effort, it's important to know what you're aiming for. It isn't always easy but it helps to weigh up your priorities and decide what can be achieved with the time and resources available to you.

• The Integrity Institute released a best practices guide last year, on defining and achieving success, with practical advice on how to build a team, create product interventions, and work with external stakeholders.
• This recent panel discussion from the Prosocial Design Network outlines how design can contribute to election integrity and goes deep into how, for example, rate limits can help democratise online spaces.
• The Kofi Annan Commission on Elections and Democracy in the Digital Age created a report on elections and democracy in 2020, with several practical suggestions for platforms.
• All Tech is Human's Tech & Democracy report is a great example of cross-organisational collaboration and recommendations.

Don't do it alone

Although great platform design and automation can help reduce issues at scale, one important theme in all of the best practices guides is the need for people: investment in human review and moderation in local languages, and collaboration with people on the ground who can give important cultural context to elections.

• Anchor Change's Playbook outlines how to create an Election Coalition. The Integrity Institute also put together a compendium of Civil Society groups.
• Network with industry peers — the oft-mentioned Integrity Institute, TSPA, and All Tech Is Human are great groups to learn from others.

Learn from what's already happened

As elections happen around the world, organisations are putting together reports on how they've gone. Two new reports just came out that do just this:

• Tech Global Institute's report on How Social Media Influencers are Scaling Surrogate Political Campaigns in Global Majority Elections discusses the complex ways that political content can influence voters, and makes recommendations for platforms around policy, human moderation, and transparency.
• Also newly released from Meta's Oversight Board — Content Moderation in an Election Year: Key Lessons for Industry — outlines human rights frameworks, emphasising that platforms must weigh decisions against the potential for real-world harm. They give many specific examples of how Meta handled moderation decisions, and make recommendations for improvements that can be applied to most platforms.

Why I'm ditching AI for pen and paper

For all the hype abut generative AI’s boost to productivity, I’ve found that many AI features create more busy work, rather than less. Recently, I've tried out two AI tools that promised to help with some of my most common work tasks: meetings and writing. Here's what I found:

Meeting note-taking and summarising

First, I tried out AI note-taking service, Fellow. After hours of sitting in meetings, it can be hard to stay focused and remember what happened, so my hope was that Fellow would make sure I didn’t miss anything.

Fellow did a good job of summarising meetings, then fed the notes and action items into their web interface. However, it took a decent amount of time to review, sort, and assign the notes and tasks that were created. I found this process even more tedious than creating the notes myself to begin with.

The act of physically writing things down and deciding what is important enough to be put on my task list is what helps me remember. I’ve noticed this in myself naturally, but Scientific American recently highlighted a study that shows that physical writing helps learning and memory by using many more parts of the brain than typing and forcing me to prioritise what I write down. Information architect consultant Jorge Arango sums this up well:

The point of personal knowledge management isn’t capturing and managing information; it’s living better by thinking better. Notes are a medium for thinking, not its replacement. (Another way to put it: the person with the most notes doesn’t “win” at the end.)

I’ll stick with my tried-and-true method: I write down physical notes during meetings, and then copy out the to-do tasks in a daily list. (I use a traveler's notebook with one book for notes and one for tasks).

Writing

My other most-dreaded work problem is writer’s block. I love to write, but it can often take me a long time to get started, and even longer to organise my thoughts. Techdirt's Mike Masnick wrote about how he uses AI as a virtual editor, to help him brainstorm and critique articles. He feeds Lex a specific key to judge his articles, and then uses the response to iterate on an article.

I tried it out on a rough draft of this article, which, to be honest, was completely half-baked. Mike’s scorecard gave me mostly 1s (which didn’t surprise me), and I suppose it was helpful to get harsh feedback early in the writing process.

I do really love Lex to write in, and especially liked the ability to choose different LLMs to critique the article, as each gave slightly different advice which helped me to keep plugging away at editing. It’s not hugely different from Google Docs, but was a small quality-of-life upgrade that I’ll keep using. Overall, I’d say that the LLM-generated suggestions were vaguely helpful in forcing me to organise my thoughts better, but certainly weren’t game-changing.

The reason why productivity hacks feel so appealing is that they promise to help you do more in less time. However, it’s the time itself that I find most valuable. The most helpful technique to improve my writing was to simply stare out the window, let my thoughts wander, and then re-read what I wrote a few minutes later with fresh eyes. Some critical parts of this piece came to me while in the shower (which is an actual thing, according to Headspace).

Ironically, experimenting with AI tools did improve my productivity, but only because they renewed my commitment to old-school analog techniques.

If you're using AI to help with your personal work in T&S, I'd love to hear from you. What’s working well?

Also worth reading

Was There A Trojan Horse Hidden In Section 230 All Along That Could Enable Adversarial Interoperability? (TechDirt)
Why? Just when we thought we'd already picked Section 230 to death, it turns out that an overlooked typo could change everything.

Online Service Providers Should Prepare For Tough New Laws In Australia (Tremau)
Why? Australia seems to be ramping up their safety regulation enforcement - here's what you should know.

Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile (NIST)
Why? This resource provides frameworks and case studies to help organizations manage AI risk.

Bonus: Alice appearances

Trust & Safety Masterclass (ElevateCX/ Vimeo)
I spoke with Denise Wilson (formerly of Twitter, Instacart and CashApp), Noor Salama of Feeld (formerly Meta), and Sarah Hatter (Elevate CX) about tips, tricks, and trends in T&S (say that five times fast).

Let's start with policy (Safety is sexy podcast)
My friend Matt Soeth started a T&S podcast, and I was his very first guest! We talk about all things policy, which is one of my favourite subjects.