How safe are today’s MLLMs?

Read the latest research

Resources

Featured Resources

Is the Safest AI Response No Response?

Appen’s latest research reveals Claude Sonnet 3.5 resisted adversarial prompting by refusing more often. Should benchmarks reward silence instead of penalizing it?

How Cohere Scaled Preference-Based Fine-Tuning for Enterprise LLMs

Discover how Cohere partnered with Appen to scale high-quality supervised fine tuning and LLM evaluation with real-time annotation.

Pre-Labeled Datasets

Accelerate your AI projects with licensable datasets

Browse our extensive catalog of over 270 audio, image, video and text datasets in over 80 languages. Our pre-labeled datasets are available immediately so you can get started right away.

Audio Data for AI Models: End-to-End Solutions from Collection to Deployment

End-to-end audio data for voice AI. Collect, transcribe, and validate speech across 500+ languages with Appen’s global network.

Blog

Is the Safest AI Response No Response?

LLM Hallucinations: Why Models Make Mistakes & How to Fix Them

How Retail AI is Transforming Shopping Experiences

ACL 2025: 5 Trends Shaping the Future of LLMs

How Krippendorff’s Alpha Improves Data Reliability

Old Is New Again: How Rubrics and Fine-Tuning Work Together in LLM Evaluation

Preserving Cultural Nuance in AI: Beyond Translation

AI Circle x Appen: Experts in the Loop

Case Studies

How Cohere Scaled Preference-Based Fine-Tuning for Enterprise LLMs

Enhancing an AI Video Description Generator with Human Validation

Accelerating Music Generation for Leading AI Platform

Rapid-Sprint LLM Evaluation & A/B Testing in Multiple Complex Domains

Improving Multilingual LLM Performance with Supervised Fine-Tuning

Training an LLM Image Generator for Graphic Design in 20+ Languages

How ReflexAI Empowers Veterans with AI Mental Health Support

How Microsoft is Advancing Equitable Knowledge Access with AI Translation

White Papers, Reports and eBooks

Audio Data for AI Models: End-to-End Solutions from Collection to Deployment

Red Teaming Multimodal Language Models: Evaluating Harm Across Prompt Modalities and Models

Retail AI: the Future of Shopping

Multilingual LLM Translation: Evaluating Cultural Nuance in Generative AI

Adversarial Prompting: Benchmarking Safety in Large Language Models

AI Safety: Building Responsible and Reliable System

Mastering Large Language Models: A Deep Dive for AI Leaders

2024 State of AI Report: Navigate the Evolving AI Landscape

More Resources

Press Releases

Data Sheets

Webinars