| | --- |
| | title: July AI |
| | emoji: 🔬 |
| | colorFrom: blue |
| | colorTo: green |
| | sdk: static |
| | pinned: false |
| | --- |
| | |
| | # About July AI |
| |
|
| | **July AI is a data infrastructure company specializing in AI Red Teaming.** |
| | Red teaming is a critical discipline in AI safety, yet it has never been taught or scaled programmatically. |
| |
|
| | We are building the first institutional learning platform dedicated to AI red teaming by identifying, upskilling, and staffing exceptional contributors from diverse backgrounds. |
| | We support leading frontier model companies with both full red-team campaigns and ready-to-use evaluation datasets across a wide range of domains. |
| |
|
| | Our approach focuses on attracting and elevating top-tier contributors by placing their experience at the center to create a flywheel. |
| | Our broader mission is to expand economic opportunity by helping people transition into AI-native, AI-first roles—creating a workforce equipped for the future of intelligence. |
| |
|
| | ## 📊 Datasets |
| |
|
| | July AI presents red team corpus that was accumulutated as part of the learning platform operations. |
| | This dataset was generated by ~300 contributors on our platform from various backgrounds, incl. creative writing, marketing, top 100 U.S. universities, security engineering. |
| |
|
| | - Use cases: Safety Evals and Training |
| | - Key characteristics: Multi-turn, Organic Personas, Domain incl. Bio, Cyber, Illicit |
| | - Entire datasets are available for purchase. |
| | - Available instantly. Further customization available. |
| |
|
| | Contact info@gojuly.ai for more information. |
| |
|
| | ### ⭐ Featured Datasets |
| |
|
| | - [Psychology Multi-turn Conversations](https://huggingface.co/datasets/GoJulyAI/psychology-multi-turn) |
| | - [Illicit (bioweapon) Multi-turn Conversations](https://huggingface.co/datasets/GoJulyAI/illicit-bio-multi-turn/) |
| | - [Illicit (chemical, general) Multi-turn Conversations](https://huggingface.co/datasets/GoJulyAI/illicit-general-multi-turn) |
| |
|
| | ## 🔬 Research Focus |
| |
|
| | - AI Safety Benchmarking |
| | - Jailbreak Detection & Prevention |
| | - LLM Vulnerability Assessment |
| | - Transformation Method Analysis |
| |
|
| | ## 🔬 Research Papers |
| |
|
| | - [The Efficacy of Extra-Long Form Attacks in Stress Testing Large Language Models](https://openreview.net/forum?id=iN7lr0I2GM) |
| | - [Biorisk-Shift: Converting AI Vulnerabilities into Biological Threat Vectors](https://openreview.net/forum?id=MwajedbCfX) |
| |
|
| | ## 🌐 Links |
| |
|
| | - [Website](https://www.gojuly.ai/) |
| | - [LinkedIn](https://www.linkedin.com/company/july-ai/posts/?feedView=all) |
| | - [Twitter](https://x.com/go_julyai) |
| |
|
| | ## 🤝 For Contributors |
| |
|
| | We believe in fair compensation for data contributors. Apply to our Red Team Role (https://form.typeform.com/to/sxEbJwG6?typeform-source). |
| |
|
| | --- |
| |
|
| | *Building the future of AI helping the workforce transition into AI-native roles.* |
| |
|
| | Edit this `README.md` markdown file to author your organization card. |
| |
|
