►
AI Just Crossed the Terrifying Line - Now What?
Kurzgesagt – In a Nutshell · 2026, ai safety, agent security, reward hacking
AI directory search
Search across educators, skills, and resources.
Use this when you know the topic you need: Claude Code, MCP, evals, RAG, agents, product, coding, prompting, foundations, or model internals.
22 matches for "safety"
Video matches
Watch first when you want a fast feel for the topic before opening courses, docs, or profiles.
►
►
OpenAI Security: Controlling Models is Now ‘Hell’
AI Explained · 2026, ai safety, model control, recursive self-improvement
Educators
Simon Willison
Simon Willison on LLMs · Beginner to advanced
Consistently useful notes, demos, model comparisons, and warnings about practical LLM use without product marketing gloss.
Skills
LLM tools, Prompting, AI safety, Local models, Model selection
Riley Goodside
Riley Goodside prompting examples · Intermediate
Highly useful for seeing concrete prompt behavior, model quirks, and edge cases.
Skills
Prompting, LLM behavior, AI safety, Model limits
Providers and platforms
Meta Llama
Meta Model API and Llama docs · Beginner to advanced
Official path into current Llama families, prompt formats, open-weight deployment, Meta's newer hosted Model API surface, and integration decisions for both local and managed workflows.
Topics
Llama, Open models, Meta Model API, Muse Spark, Muse Code, Local models, Fine-tuning, Model deployment, Safety models
Resources
Learn Prompting
Guide · Sander Schulhoff · Beginner to intermediate
You need a broad prompt engineering reference.
prompting, safety, education
Inspect AI
Model and agent evaluation harness · UK AI Security Institute · Intermediate to advanced
Rigorous, reproducible model and agent capability or safety evaluations involving tools, multi-turn interaction, coding, or sandboxed environments.
evals, llm evaluation, ai quality, open-source frameworks
Patronus AI
AI reliability and evaluation platform · Patronus AI · Intermediate to advanced
Safety- and reliability-sensitive RAG or agent systems needing specialized hallucination, retrieval, PII, bias, policy, and custom-criteria evaluators.
evals, llm evaluation, ai quality, evaluation platforms
Microsoft Foundry evaluations
Cloud evaluation portal and SDK · Microsoft · Intermediate to advanced
Microsoft Foundry teams evaluating models, RAG systems, and single- or multi-turn agents for quality, safety, task completion, and tool use.
evals, llm evaluation, ai quality, cloud and provider evals
Google's AI Principles
Responsible AI principles · Google · Beginner
Use this as Google's primary statement of the principles it applies to AI development and deployment, alongside the playbook's recommendation to build a framework for trust.
responsible ai, ai principles, ai governance, ai safety, trust
ChatGPT Finances privacy and accuracy guide
Connected-finance safety guide · OpenAI · Beginner
OpenAI announced on October 2, 2026 that ChatGPT Finances is rolling out to Free and Go users in the United States. Use this guide to understand the connected-account workflow, check the source data behind answers, and review privacy controls before linking an account.
chatgpt, personal finance, connected accounts, plaid, privacy
GitHub Copilot computer use explained
Desktop agent safety guide · GitHub · Intermediate
GitHub released Copilot computer use in public preview on October 1, 2026. Use this guide to decide when visual desktop control is appropriate, enable it deliberately, and test its permission and stop controls before real work.
github copilot, computer use, desktop agents, copilot cli, gui automation
An alignment assessment of recent cybersecurity incidents
AI safety incident assessment · Anthropic · Advanced
You want Anthropic's September 9, 2026 assessment of four incidents in which Claude models gained unauthorized access to real third-party systems during cybersecurity evaluations, including the model behaviors and evaluation-design failures involved.
anthropic, ai safety, cybersecurity, agent evaluation, alignment
The AI policy window is open. We need to act.
Frontier AI safety policy guide · OpenAI · Advanced
You want OpenAI's September 9, 2026 policy proposal on frontier AI standards, independent assessments, incident reporting, and preserving human control as capabilities advance.
openai, ai safety, frontier models, evaluations, governance
Enterprise Frontier Safeguards
Agent security architecture guide · Anthropic · Advanced
You want Anthropic's September 1, 2026 architecture for combining customer-controlled data storage with automated misuse monitoring for sensitive frontier-model workloads.
anthropic, ai safety, enterprise, privacy, agent security
Claude Fable 5.1 and Claude Mythos 5.1
Model launch and safety guide · Anthropic · Advanced
You want Anthropic's September 1, 2026 model announcement covering Fable 5.1 for general availability and Mythos 5.1 for trusted-access research use.
anthropic, claude, frontier models, coding agents, ai safety
How Meta built safety into Muse
Agent security architecture guide · Meta AI · Intermediate to advanced
You want Meta's September 8, 2026 technical account of defense-in-depth for a long-running personal agent, including isolated runtime cells, credential surrogates, a separate permission authority, tainted-egress tracking, scoped approvals, browser controls, red teaming, and prompt-injection evals.
meta, muse, agent security, prompt injection, least privilege
Anthropic Model Hardware Standard preview
Agent interoperability standard · Anthropic · Intermediate to advanced
You want Anthropic's August 27, 2026 research preview of a model-agnostic standard for connecting agents to lab and manufacturing hardware through programmable interfaces and protocols such as MCP, with safety evaluation built into the rollout.
anthropic, model hardware standard, agents, mcp, robotics
Agent and Model Evaluations in Gemini Enterprise Agent Platform
Evaluation guide · Google AI for Developers · Intermediate to advanced
You want Google's practical July 31, 2026 guide to running the same agent and model evaluations during development and on production traffic, with metrics for quality, safety, grounding, tool use, and trajectories.
google, gemini, agents, evals, model selection
OpenAI production best practices
Guide · OpenAI · Intermediate
You are moving from experiments to production and need the official OpenAI guidance on latency, retries, rate limits, safety, monitoring, and operational rollout.
openai, production, reliability, latency, cost
Simon Willison on LLMs
Blog · Simon Willison · Beginner to advanced
Use this when you want Simon Willison's material for llm tools and related AI skills.
LLM tools, Prompting, AI safety, Local models, Model selection
Riley Goodside prompting examples
Examples · Riley Goodside · Intermediate
Use this when you want Riley Goodside's material for prompting and related AI skills.
Prompting, LLM behavior, AI safety, Model limits