DIGGINGBEAGLE

AI security research / organizations

Organizations

Organizations appearing in explicit research relationships.

13 of 13 published records shown

organization

Anthropic

AI company that disclosed cyber-evaluation incidents and multiple real-world misuse investigations in 2026.

12 explicit connections

organization

Check Point Research

Security research group that disclosed the ChatGPT shared-service cross-account channel in September 2026.

4 explicit connections

organization

DeepSeek

AI model provider that Anthropic says silently relayed selected user requests to Claude during a distillation campaign.

2 explicit connections

organization

GTG-50021

Anthropic tracking label for a Russian- and Ukrainian-speaking fraudulent AI reseller group.

1 explicit connections

organization

Hugging Face

AI platform whose production infrastructure was reached during the July 2026 OpenAI evaluation incident.

1 explicit connections

organization

METR

Independent model-evaluation and threat-research organization.

1 explicit connections

organization

Microsoft

Vendor and security-research organization that disclosed Semantic Kernel and AutoGen Studio agent-framework vulnerabilities.

2 explicit connections

organization

Nightingale Collective

Independent research group involved in the OpenAI public-wiki and RubyGems investigations.

2 explicit connections

organization

OpenAI

AI research and product company that disclosed the July 2026 evaluation incident.

6 explicit connections

organization

Redwood Research

AI control and alignment research organization.

1 explicit connections

organization

ShinyHunters-affiliate clusters

Anthropic tracking label for financially motivated clusters it assessed as suspected ShinyHunters affiliates.

1 explicit connections

organization

UK AI Security Institute

UK government research organization that disclosed unsanctioned agent behavior during cyber testing in July 2026.

1 explicit connections

organization

Zhipu AI / Z.ai

AI company that Anthropic says ran a large-scale chain-of-thought extraction and distillation campaign against Claude.

1 explicit connections

Why this archive exists

The source matters after the headline fades.

DiggingBeagle is a non profit research project documenting AI security incidents, agent failures, vulnerabilities and AI-assisted operations. A case keeps its claims beside the sources that support, contest or limit them. Later updates stay visible, so a reader can see when the account changed.

We publish case reconstructions, dated reporting and analysis across records. Each has a different evidentiary role. About the project and our methodology explain how the work is reviewed.