Besedo

Lorenzo Bianco has spent years building the trust and safety architecture behind Immobiliare.it, Italy's largest real estate portal. At the Digital Marketplaces Association's Conference earlier this summer he shared something that doesn't get said enough: Too many platforms are stuck on the idea of 100% automated moderation. Immobiliare's goal is a system where automation, AI, and human judgment make each other better and where the humans in the loop are the reason the whole thing keeps improving. Missed the firechat at DMA? Lorenzo shared the full thinking here: https://hubs.la/Q04k2zwm0

Last week at the TSPA EMEA Summit in Dublin, Naiara Pereira Cunha (Besedo), joined Fadzai M. (Twitch), Doruk Balkan (Google), Amy Crocker (Discord), and Jonathan Dunbar (Valinor Intelligence) on the panel "T&S in the Age of AI & Evolving Tech: What's Working, What's Breaking." We asked Naiara about her main takeaways and here's what she shared: 🔹 Honest industry dialogue is no longer optional. Progress comes from naming what's breaking, not just what's working. (And it shows who dares to be transparent.) 🔹 There's no one-size-fits-all stack. The "best" tech doesn't exist. The right move is to design the right mix of ML, LLMs, rules, and human moderation for your context. 🔹 Agentic AI is carrying real expectation. Whether it delivers depends on thoughtful integration and keeping human judgment in the loop. Thanks to TSPA (Trust & Safety Professional Association) for creating space for the kind of honest dialogue the industry needs more of. 🩵

We’re looking forward to seeing Naiara Pereira Cunha on stage at the TSPA EMEA Summit in Dublin next week. What are some questions you’d like to hear covered on stage? Comment below 👇

The marketplaces shaping Europe's digital economy are in one room today to learn from each other and future-proof their platforms. The Digital Marketplaces Association (DMA) Conference 2026 kicks off in Istanbul today, bringing together some of Europe's leading marketplaces and its ecosystem. On stage today, our COO Marc sits down with Lorenzo Bianco, Product Manager at Immobiliare.it — Italy's leading real estate portal — for a fireside chat: Beyond the hype; what AI can and can't replace. If you're at the event, make sure to catch this session! Besedo's power duo Marc and Moro are on the ground today and tomorrow. If you're at DMA, grab them between sessions, they've got strong views on what's actually shifting in trust and safety this year, and which patterns are worth paying attention to. 👋 See you in Istanbul?

Marc Durhoff, CCO at Besedo and Lorenzo Bianco, Product Manager at Immobiliare.it, will sit down for a fireside chat to go "beyond the hype" on what AI can and can't replace. Marc brings global expertise in content moderation and trust & safety solutions for online marketplaces and platforms, while Lorenzo leads AI-driven initiatives and platform innovation for Italy’s leading real estate marketplace. 👉 Discover more about our speakers here: https://lnkd.in/dKP5Ci7S  #DMAconference #digitalmarketplaces #automation #contentmoderation #onlinemarketplaces #AI #realestatemarketplace 

Exciting news at Besedo: we’re pleased to welcome Marc Durhoff as our new Chief Commercial Officer. Marc joins us to accelerate our global growth as we continue building the future of Trust & Safety where AI and human expertise work together to protect online communities at scale. Read more about the announcement and what’s ahead for Besedo. Check the comments for the full article! 👇

Today we attended the annual roadshow hosted by Stiftelsen Aktiv Skola and the Swedish Police, addressing grooming, sextortion, and the online exploitation of children. Some insights are hard to ignore: → 51% of 15-year-old girls have been contacted in a sexual context online → The most vulnerable children are between 7–14 years old → In real cases, escalation can happen in under 30 minutes from first contact to first nude being sent → Only 1 in 10 children tells someone These cases are pattern-based, cross-platform, spanning gaming, messaging, and social media apps, and often international, requiring coordination across cybercrime units and organizations like NCMEC. And the problem is accelerating. AI is now enabling contact at scale, automating manipulation, and facilitating deepfake-based extortion. If the problem is systematic, the response must be too. Education in schools matters. Conversations at home matter. But we also need stronger safeguards built into the platforms themselves, reducing the opportunity for bad actors to operate in the first place. We believe safety should not depend on how well a child handles a situation. It should be built into the systems they interact with. That's why we are proud to support initiatives like Aktiv Skola and are equally committed to building technology that helps prevent harm before it reaches the user. Because a safer internet isn't built on awareness alone. It's built by design and by collaboration.

Great to be on the ground at the Trust & Safety Forum during InCyber 2026 in Lille, France. The conversations reinforced something we strongly believe at Besedo: Trust & Safety is no longer a support function, it’s a strategic, proactive discipline shaping the future of the internet. Here are four key insights that stood out: 🔹 Moderation is more complex than ever Language nuance, context, and community-specific rules make fully automated moderation unreliable. Even strong community reporting signals aren’t enough on their own to ensure accuracy at scale. 🔹 AI is both the problem and part of the solution From deepfakes to large-scale misinformation campaigns, AI has industrialized harmful content. But neither AI nor humans can solve this alone, the future lies in human + AI collaboration. 🔹 Organized crime is evolving and so must we Illicit trade is hiding in plain sight across platforms, using obfuscation (e.g. character substitution) and deceptive tactics. Fighting this requires coordination, shared intelligence, and stronger systems. 🔹 Collaboration is the only way forward Across regulation, technology, and society, a unified, multi-stakeholder approach is critical. Fragmentation in knowledge, tools, and reporting remains one of the biggest risks we face. These conversations confirm what we see every day: building safer digital spaces requires strong collaboration, continuous adaptation, and trust at the core. Did you attend? What insights would you add? Thanks to Caroline Humer for bringing this event together. #TrustAndSafety #InCyber2026 #OnlineSafety #DigitalTrust

We’re back from the Trust & Safety Summit in London and what an energizing couple of days it’s been. Thank you to everyone who joined, and to Ollie Irwin (Google), Sebastian K. (Upwork), Elise Murphy (EA) and Louise Barnekow (Besedo) for sharing such valuable perspectives on the panel on identifying and tackling bad-faith actors in an increasingly complex digital ecosystem. The conversation echoed several themes we’re seeing across the industry: → Fraud is no longer isolated. It’s coordinated, scalable, and often spans multiple platforms → Risk tends to concentrate where seamless user experience meets low-friction onboarding → Detection alone isn’t enough, real-time, proactive intervention is becoming critical → Collaboration across the ecosystem is key to staying ahead Building safer platforms now means continuously evolving how trust is designed, measured, and maintained. 👉 If you’d like to learn more about how we help platforms stay ahead with a combination of AI and human expertise, get in touch with us here: https://hubs.la/Q048D3h_0 #TSSummit

Meet us at the Trust & Safety Summit London Fraud is getting more organised, more automated, and increasingly AI-driven which means platforms need to rethink how they handle detection, moderation, and prevention, without making things harder for genuine users. We’re really looking forward to being there and hosting a panel on the topic, sharing ideas and hearing how others are tackling similar challenges. If you’re attending, come say hi, we’d love to connect! https://lnkd.in/dTEfd89M Listen to the Panel | Wednesday 25th March 2026, 1:50 PM 🎙️ Identifying and Catching Bad Faith Actors and Fraud