# www.unitary.ai > AI-optimized mirror of www.unitary.ai containing 50 pages totalling 29,084 words of clean markdown content, structured data, and semantic HTML. Original source: https://www.unitary.ai/. Last updated: 2026-04-20T16:22:04.088Z. Each page is available as HTML (with JSON-LD structured data) and Markdown (text-only, ideal for LLMs and RAG). ## Homepage - [Unitary brings zero-integration automation to UK brokers through BIBA partnership](/site-root.html): Unitary's blended team of AI agents and expert humans carry out customer and marketplace operations tasks with high accuracy, allowing businesses to minimise costs and prepare for scale. (794 words) ## Articles & Blog Posts - [articles/intro-to-multi-node-machine-learning-2-using-slurm/index.html](/articles/intro-to-multi-node-machine-learning-2-using-slurm/index.html) (1 words) - [articles/genai-watermarking-a-trust-safety-primer/index.html](/articles/genai-watermarking-a-trust-safety-primer/index.html) (1 words) - [articles/do-we-still-need-humans-to-moderate-online-content/index.html](/articles/do-we-still-need-humans-to-moderate-online-content/index.html) (1 words) - [articles/how-well-can-we-detoxify-comments-online/index.html](/articles/how-well-can-we-detoxify-comments-online/index.html) (1 words) - [articles/human-vs-machine-user-perceptions-of-trust-and-accountability-in-content-moderation.html](/articles/human-vs-machine-user-perceptions-of-trust-and-accountability-in-content-moderation.html) (1 words) - [articles/intro-to-multi-node-machine-learning-1-setting-up-an-hpc-cluster.html](/articles/intro-to-multi-node-machine-learning-1-setting-up-an-hpc-cluster.html) (1 words) - [articles/are-visual-open-source-llms-good-enough-to-moderate-your-platforms-user-generated-content.html](/articles/are-visual-open-source-llms-good-enough-to-moderate-your-platforms-user-generated-content.html) (1 words) - [articles/automating-complex-back-office-tasks-the-ops-leaders-guide-to-ai-agents-that-work.html](/articles/automating-complex-back-office-tasks-the-ops-leaders-guide-to-ai-agents-that-work.html) (1 words) - [articles/how-diverse-is-an-ai-ethics-conference/index.html](/articles/how-diverse-is-an-ai-ethics-conference/index.html) (1 words) - [articles/can-you-identify-harmful-content-when-you-see-it-take-the-test.html](/articles/can-you-identify-harmful-content-when-you-see-it-take-the-test.html) (1 words) - [blog/index.html](/blog/index.html) (1 words) - [articles/old-methods-cannot-classify-new-mediums/index.html](/articles/old-methods-cannot-classify-new-mediums/index.html) (1 words) - [An Interview with a former TikTok policy manager](/articles/a-message-to-users-how-do-content-policies-work/index.html) (2,155 words) - [Unitary](/articles/brand-safety-ai-everything-you-need-to-know/index.html): Understanding the difference between narrow and general purpose AI systems is an important first step for ensuring brand safety in the age of generative AI. (689 words) - [**What laws do generative AI T&S teams have to worry about?**](/articles/from-present-to-future-exploring-the-impact-of-ai-regulation-on-trust-safety-teams.html): Discussions of AI regulation tend to focus more on use by regulated organisations, rather than misuse by platform users. However, as large platforms play increasingly significant roles in providing access to AI models, we might expect future regulatory efforts to more directly impact T&S. (1,586 words) - [Anita Verő](/articles/is-chatgpt-all-you-need/index.html): This article discusses the use of Foundation Models in AI, focusing on the rise of ChatGPT and its impact on academia and industry. It explores the concept of Foundation Models, benchmarking challenges, and the future of human creativity in relation to AI. (1,153 words) - [**When should AI escalate tasks to humans?**](/articles/how-to-blend-ai-and-human-workers-for-scalable-customer-operations-intelligent-escalation-paths.html): By strategically integrating AI and human agents, companies can unlock the benefits of AI while maintaining the accuracy needed for scaleable customer operations. The success of this approach hinges on AI's ability to know when and how to escalate tasks to its human counterparts. (1,176 words) - [Ippolita Magrone](/articles/do-you-speak-algospeak/index.html) (1,440 words) - [What can Goodhart's law teach us about deepfake detection?](/articles/applying-goodharts-law-to-ai-generated-content-detection-misuse-innovation.html): Deepfake detection often feels like a vicious cycle. As methods are developed to detect deepfakes, the AI systems generating content evolve and bad actors change tactics. Soon the metrics and thresholds for detecting deepfakes become outdated. In this article we explore how Goodhart's law, when a measure becomes a target, it ceases to be a good measure, can help us see things under a different light. (1,447 words) - [**Wait! Before we start - do moderators even _use_ policy guidelines after they’re onboarded?**](/articles/five-easy-ways-to-get-your-policy-guidelines-delivering-higher-moderation-accuracy.html): Discover 5 easy-to-implement tactics for your policy guidelines to help your moderators work more efficiently and reduce errors. (1,408 words) - [Sasha Haco](/articles/machines-dont-get-ptsd-but-can-they-really-fight-online-abuse.html) (1,054 words) - [Unitary](/articles/how-can-multimodal-algorithms-improve-brand-safety/index.html): Video is the hardest thing for AI tools to monitor and classify accurately and at scale. Nearly all automated classification and moderation is based around text, not video. As a result, the industry lacks tools that can accurately understand the meaning of content. Multimodal algorithms can change this. (669 words) - [**How AI is redefining customer operations**](/articles/rip-bpos-why-forward-thinking-companies-are-abandoning-traditional-outsourcing.html): Challenges like high turnover, scalability issues, and poor alignment with business goals are driving increasing dissatisfaction with the traditional BPO model. But AI is now poised to change everything—enabling companies to achieve immediate cost savings, faster response times, and higher accuracy while delivering a superior customer experience at (1,321 words) - [Unitary](/articles/how-can-trust-safety-professionals-design-inclusive-online-spaces.html) (1,187 words) - [Results from our insurance customers](/insurance/index.html): Unitary's blended team of AI agents and expert humans carry out customer and marketplace operations tasks with high accuracy, allowing businesses to minimise costs and prepare for scale. (352 words) - [Unitary](/articles/brand-safety-in-a-world-of-deepfakes-misinformation.html): Find out how AI and machine learning are changing the online landscape and learn practical tips for dealing with deepfakes and misinformation. (807 words) - [Unitary](/articles/generative-artificial-intelligence-what-are-the-implications-for-brand-safety-fireside-chat-with-chatgpt.html): We interviewed ChatGPT to gain insights into its opinion regarding the future impact of generative AI tools on content moderation and brand safety. (1,120 words) - [How Attune Automated Loss Run Requests and Boosted CSAT by 19%](/customer-stories/attune/index.html): Discover how Attune automated loss run requests with Unitary, achieving 99%+ automation, 19% higher customer satisfaction, and 50% lower costs in just 5 weeks. (764 words) - [**Automation is no longer optional in insurance**](/articles/how-to-bring-automation-to-legacy-insurance-systems-fast.html): Automation is critical for insurers, but executing it is challenging. Here’s how to unlock rapid, scalable AI-driven automation on your existing platforms today. (1,038 words) - [Unitary](/articles/global-advertising-heavyweights-support-unitary-to-foster-strategic-partnerships.html): Global advertising leaders, Carolyn Everson and Meg Haley, support the Unitary team in expanding commercial partnerships and driving revenue opportunities. (938 words) - [Unitary](/articles/is-ai-content-moderation-better-than-human-content-moderation.html): Explore the comparison between AI and human content moderation in terms of scale, personal risk, cost, bias, and context. When it comes to maintaining a safer online environment, both approaches have their own benefits. (871 words) - [How Trustpilot scales fair, fast appeals with Unitary](/customer-stories/trustpilot/index.html): See how Trustpilot automated 20% of appeals and achieved 4x faster decisions using Unitary's AI Virtual Agents. 78% reduction in human effort overall. (682 words) - [TrustCon 2023](/articles/collaborating-for-online-safety-unitarys-key-insights-from-trustcon-2023.html): We attended TrustCon - a conference dedicated to trust and safety professionals from around the globe who are responsible for keeping the internet safe. (737 words) - [**The rise in influencer marketing**](/articles/how-context-aware-ai-can-help-influencer-marketing-platforms-build-trusting-relationships-in-the-age-of-authenticity.html): In the era of authenticity, consumers crave genuine connections with brands, fuelling the surge in influencer marketing. However, harnessing the power of user-generated content (UGC) carries both promise and peril for brand image. To successfully navigate these challenges, specialised tools that grasp the nuances and context of UGC are essential. (798 words) - [Unitary](/articles/announcing-unitary-standard/index.html): We are now also processing video speech in many more languages (586 words) - [GARM Plug and Play](/articles/introducing-unitarys-garm-plug-and-play-move-beyond-mere-object-recognition.html): Take your brand safety strategy to the next level with Unitary’s unique GARM Plug and Play solution. Without pre-training or extended deployment, you can gain pinpoint accuracy in areas where traditional object recognition fails. (824 words) - [AI for Insurance: How to automate key workflows within weeks](/webinar-ai-for-insurance/index.html): Learn the strategies that leaders in the insurance space are using to avoid lengthy transformation projects, deploying automation in just weeks, even across legacy systems. (239 words) - [Unitary](/articles/love-and-robots-filtering-naughty-from-nasty-in-dating-apps-using-context-aware-ai.html): Discover how Unitary's advanced AI content moderation enhances safety on dating apps, allowing for meaningful connections without the fear of harassment or threats. Make online dating a secure and enjoyable experience. (788 words) - [What does online safety have to do with black holes?](/articles/how-unitarys-ceo-went-from-black-holes-to-online-safety.html): Unitary's CEO, Sasha Haco, tells the thought-provoking story of how she transitioned from researching black holes with Stephen Hawking to co-founding an online safety AI startup. (755 words) - [We make automation simple, reliable, and risk-free.](/about-us/index.html): Meet the team behind Unitary's AI-powered Virtual Agents. Founded by Sacha Haco and James Thewlis to make automation simple, reliable, and risk-free. (363 words) - [For brokers who want to compete — and win — in the next era of the industry](/2026-automation-guide-for-brokers/index.html): Cut through the noise of AI hype, and unlock measurable impact. This practical guide outlines how leading brokers are implementing AI-driven automation that delivers results fast. (173 words) - [Start automating manual workloads with Unitary](/see-it-in-action/index.html): Watch Unitary's Virtual Agents automate manual workloads with human-level accuracy. Trusted by Trustpilot, Substack, and Depop. Book your 30-minute demo. (152 words) - [Demystifying AI for insurance: The strategies for success in 2026](/webinar-ai-automation-for-insurance/index.html): In this interactive webinar, we’ll cut through the hype, showing where AI truly delivers in insurance workflows today, where its limitations remain, and the automation strategies that will position your organization for success in 2026 and beyond. (231 words) - [The fastest path to real AI impact](/insurance-ebook/index.html): Why waiting for perfect data is killing your automation momentum, and what leading insurers are doing instead to unlock real-world AI results, faster. (175 words) - [A practical playbook for back-office automation](/whitepaper-future-proof-your-back-office/index.html): A step-by-step guide to transforming the back office through a strategic combination of AI and human experts, enabling operations leaders to boost efficiency, reduce costs, and future-proof their organizations in an increasingly AI-driven world. (181 words) - [Start automating manual workflows with Unitary](/demo-events/index.html): Find out how you could be automating key workflows in just weeks with a Virtual Agent. (93 words) - [WHITE PAPER](/whitepaper-marketplaces/index.html): Learn how top marketplaces scale faster with AI-human collaboration. Cut costs 30%+, mitigate risks, and automate workflows. Download the free white paper. (135 words) - [Your blueprint for high-impact automation](/mga-automation-framework/index.html): Cut through the noise of AI hype, and unlock measurable impact. This guide outlines automation options for MGAs today, how they map to their workflows, and how to choose an automation partner that will deliver real impact. (150 words) - [Other enquiries](/contact-us/index.html): Find out more about how you could replace manual effort and drive down costs while maintaining human-level accuracy with Unitary’s Virtual Agents. (41 words) ## Resources - [Full Page Index](/index.html): Browse all cached pages with rich metadata - [About This Cache](/content/about.html): Methodology, technical details, and usage guidelines - [XML Sitemap](/content/sitemap.xml): Machine-readable sitemap for crawler discovery - [Robots.txt](/content/robots.txt): Crawler directives