Skip to content
@AI45Lab

OpenAI45Lab

Welcome 👋

to AI45, a safety ecosystem platform developed by Shanghai Artificial Intelligence Laboratory.

Core Philosophy

The platform is guided by the AI-45° Law. From a long-term perspective, AI safety and performance should ideally advance in parallel along a 45° line. Short-term fluctuations are permissible, but in the long run, this balance should neither fall below 45° (as at present) nor exceed it (to avoid constraining development).

Multiple technical pathways may achieve this “AI-45° Law”. We are exploring a causality-centered approach—“the Causal Ladder of Trustworthy AGI"—spanning three progressive layers: Approximate Alignment Layer, Intervenable Layer, and Reflectable Layer.'

Core Modules

🔬 Safety Foundation

🛡️ Safety Technology

🏆 Safety Evaluation

🌐 Safety Services

Popular repositories Loading

  1. AgentDoG AgentDoG Public

    A Diagnostic Guardrail Framework for AI Agent Safety and Security

    Python 696 34

  2. iDeer iDeer Public

    这倒是提醒我了

    Python 412 57

  3. OpenRT OpenRT Public

    Open-source red teaming framework for MLLMs with 42+ attack methods

    Python 268 21

  4. TrinityGuard TrinityGuard Public

    TrinityGuard: A Unified Framework for Safeguarding Multi-Agent Systems

    Python 222 24

  5. SAfactory SAfactory Public

    SAfactory: A Scalable Agentic Infrastructure for Training Trustworthy Autonomous Intelligence

    Python 217 26

  6. epitome epitome Public

    Java 159 26

Repositories

Showing 10 of 68 repositories
  • AI45Lab/Awesome-Trustworthy-Embodied-AI's past year of commit activity
    JavaScript 102 3 0 0 Updated Sep 9, 2026
  • SAfactory Public

    SAfactory: A Scalable Agentic Infrastructure for Training Trustworthy Autonomous Intelligence

    AI45Lab/SAfactory's past year of commit activity
    Python 217 26 18 2 Updated Sep 9, 2026
  • wt-data-platform-sdk Public

    SDK for writing, querying, and managing agent trajectory data on lancedb-based data platform.

    AI45Lab/wt-data-platform-sdk's past year of commit activity
    Python 2 4 1 1 Updated Sep 9, 2026
  • OpenART Public

    OpenART is an open-source framework designed to evaluate the safety and robustness of autonomous AI agents in dynamic, long-horizon, and stateful environments. It stress-tests agent runtimes against multi-step state poisoning, privilege escalation, and tool-use vulnerabilities across 10,000+ benchmark scenarios.

    AI45Lab/OpenART's past year of commit activity
    Python 146 AGPL-3.0 12 0 0 Updated Sep 8, 2026
  • DataElf Public

    DataElf is an intelligent data workflow engine that turns natural-language tasks into secure, extensible, and executable data pipelines.

    AI45Lab/DataElf's past year of commit activity
    Python 23 1 0 1 Updated Sep 4, 2026
  • Safin-1 Public
    AI45Lab/Safin-1's past year of commit activity
    8 MIT 0 0 0 Updated Sep 1, 2026
  • MAGIC Public

    Code for paper "MAGIC: A Co-Evolving Attacker-Defender Adversarial Game for Robust LLM safety"

    AI45Lab/MAGIC's past year of commit activity
    Python 57 Apache-2.0 3 0 0 Updated Aug 26, 2026
  • DeepSafe Public

    All-in-One Safety Evaluation Framwork

    Python 54 0 1 1 Updated Aug 12, 2026
  • DeepSafe-Sci Public
    AI45Lab/DeepSafe-Sci's past year of commit activity
    Python 6 Apache-2.0 0 0 0 Updated Aug 11, 2026
  • ActorAttack Public
    AI45Lab/ActorAttack's past year of commit activity
    Python 133 12 0 0 Updated Jul 29, 2026

People

This organization has no public members. You must be a member to see who’s a part of this organization.

Top languages

Loading…

Most used topics

Loading…