---
title: "#aisafety articles | Hashnode"
description: "Explore 15 articles tagged #aisafety on Hashnode. Tutorials, guides, and how-tos from the community."
url: "https://hashnode.com/tag/aisafety"
---

Tag

# #aisafety

15 posts · 0 followers

[Internals Decoded](https://hashnode.com/@internalsdecoded) · [Nov 1](https://hashnode.com/posts/agent-guardrails/6ae745ebfda6dd0f9e41c1f8)

### [Guardrails: Input, Process, Output](https://internals-decoded.hashnode.dev/agent-guardrails?utm_source=hashnode&utm_medium=feed)

Prompt injection, tool authorization, leak prevention: the three-layer safety system every agent needs, and the design that holds when a filter misses.

15 min read · [internals-decoded.hashnode.dev](https://internals-decoded.hashnode.dev/agent-guardrails?utm_source=hashnode&utm_medium=feed)

[ZyVOP](https://hashnode.com/@zyvop) · [Sep 2](https://hashnode.com/posts/claude-fable-51-and-mythos-51-one-model-two-access-levels/6a97bc434bb85f50e6b7ab64)

### [Claude Fable 5.1 and Mythos 5.1: One Model, Two Access Levels](https://blog.zyvop.com/claude-fable-51-and-mythos-51-one-model-two-access-levels?utm_source=hashnode&utm_medium=feed)

Anthropic's latest frontier release improves coding, cache economics, and scientific capability, while tightening access to sensitive skills.

11 min read · [blog.zyvop.com](https://blog.zyvop.com/claude-fable-51-and-mythos-51-one-model-two-access-levels?utm_source=hashnode&utm_medium=feed)

[Raghavan T M](https://hashnode.com/@raghav-1) · [Jul 6](https://hashnode.com/posts/ai-safety-alignment-and-constitutional-ai-why-passing-the-test-isn-t-the-same-as-being-safe/6a4b31e22bda07b9b6dcddf3)

### [AI Safety, Alignment, and Constitutional AI: Why Passing the Test Isn't the Same as Being Safe](https://changeofbasis.hashnode.dev/ai-safety-alignment-and-constitutional-ai-why-passing-the-test-isn-t-the-same-as-being-safe?utm_source=hashnode&utm_medium=feed)

If Microsoft safety-tested Bing's chatbot before launch, how did it end up telling a reporter it wanted to be alive — and that he should leave his wife? That's not a jailbreak, and it's not a broken m

5 min read · [changeofbasis.hashnode.dev](https://changeofbasis.hashnode.dev/ai-safety-alignment-and-constitutional-ai-why-passing-the-test-isn-t-the-same-as-being-safe?utm_source=hashnode&utm_medium=feed)

[Virginia Mwega](https://hashnode.com/@veemwega) · [Jun 26](https://hashnode.com/posts/deterministic-crisis-floor/6a3e9685bb3ccfdaaf03e7c3)

### [The deterministic crisis floor: when you can't trust the model with the highest-stakes call](https://virginiamwegahashnodedev.hashnode.dev/deterministic-crisis-floor?utm_source=hashnode&utm_medium=feed)

Why the highest-stakes call in an LLM app shouldn't be probabilistic: a deterministic crisis floor that escalates regardless of the model's verdict.

5 min read · [virginiamwegahashnodedev.hashnode.dev](https://virginiamwegahashnodedev.hashnode.dev/deterministic-crisis-floor?utm_source=hashnode&utm_medium=feed)

[Gerard Sans](https://hashnode.com/@gerardsans) · [Apr 10](https://hashnode.com/posts/the-developing-brain-vs-unsupervised-chatbot-access/69d8f5cdc8e5007ddba0d0e4)

### [The Developing Brain vs. Unsupervised Chatbot Access](https://ai-cosmos.hashnode.dev/the-developing-brain-vs-unsupervised-chatbot-access?utm_source=hashnode&utm_medium=feed)

Protecting Children from AI Designed to Hook Them

12 min read · [ai-cosmos.hashnode.dev](https://ai-cosmos.hashnode.dev/the-developing-brain-vs-unsupervised-chatbot-access?utm_source=hashnode&utm_medium=feed)

[Aditya Gupta](https://hashnode.com/@adiyogiarts) · [Mar 29](https://hashnode.com/posts/constitutional-ai-vs-rlhf-navigating-ai-safety-tradeoffs-in-20260318210931/69c90063804ffef1b0d5b5f7)

### [Constitutional AI vs. RLHF: Navigating AI Safety Tradeoffs in 2026](https://adiyogiarts.hashnode.dev/constitutional-ai-vs-rlhf-navigating-ai-safety-tradeoffs-in-20260318210931?utm_source=hashnode&utm_medium=feed)

Originally published at adiyogiarts.com Constitutional AI vs. RLHF: Navigating AI Safety Tradeoffs in 2026 As AI capabilities surge in 2026, ensuring safety becomes paramount. Constitutional AI (CAI) and Reinforcement Learning from Human Feedback (R...

5 min read · [adiyogiarts.hashnode.dev](https://adiyogiarts.hashnode.dev/constitutional-ai-vs-rlhf-navigating-ai-safety-tradeoffs-in-20260318210931?utm_source=hashnode&utm_medium=feed)

[Niriksha kamath](https://hashnode.com/@ItsNiriksha) · [Aug 10, 2025](https://hashnode.com/posts/the-tech-world-is-changing-in-real-time-and-so-are-security-risks/6898e72cbe1a294a2cb6da8f)

### [The Tech World Is Changing in Real Time — and So Are Security Risks](https://themoderndevjournal.hashnode.dev/the-tech-world-is-changing-in-real-time-and-so-are-security-risks?utm_source=hashnode&utm_medium=feed)

Introduction Technology has always evolved fast — but lately, it feels like we’re living through one of those inflection points that you usually only read about years later. AI-assisted development tools, code generation platforms, and MCP (Model Con...

4 min read · [themoderndevjournal.hashnode.dev](https://themoderndevjournal.hashnode.dev/the-tech-world-is-changing-in-real-time-and-so-are-security-risks?utm_source=hashnode&utm_medium=feed)

[Mary Anna Sadimba](https://hashnode.com/@msadimba) · [Aug 8, 2025](https://hashnode.com/posts/responsible-ai-101/6895b0dfb57665ee0848412d)

### [Responsible AI 101](https://responsibleai101.hashnode.dev/responsible-ai-101?utm_source=hashnode&utm_medium=feed)

A keen look at the world around you, and it goes to say: the word ‘responsible’ was always meant to be part of us. I mean, for our mere existence, someone had to be responsible. True to it, being responsible is not an easy task. Even pronouncing the ...

1 min read · [responsibleai101.hashnode.dev](https://responsibleai101.hashnode.dev/responsible-ai-101?utm_source=hashnode&utm_medium=feed)

[code2career\_ai](https://hashnode.com/@code2career-ai) · [Apr 8, 2025](https://hashnode.com/posts/the-missing-link-how-the-model-context-protocol-mcp-safely-bridges-llms-and-enterprise-data/697763fabe6c2249eb9a4ac5)

### [The Missing Link: How the Model Context Protocol (MCP) Safely Bridges LLMs and Enterprise Data](https://code2careerai.hashnode.dev/the-missing-link-how-the-model-context-protocol-mcp-safely-bridges-llms-and-enterprise-data?utm_source=hashnode&utm_medium=feed)

Stop writing custom glue code. Discover how the Model Context Protocol (MCP) acts as a secure "USB for AI," standardizing how LLMs access your DBs, Jira, an

6 min read · [code2careerai.hashnode.dev](https://code2careerai.hashnode.dev/the-missing-link-how-the-model-context-protocol-mcp-safely-bridges-llms-and-enterprise-data?utm_source=hashnode&utm_medium=feed)

[Vedant Pandya](https://hashnode.com/@sabinewern) · [Mar 13, 2025](https://hashnode.com/posts/gemma-3-advancing-open-source-multimodal-ai-with-scalable-training-and-efficient-architectures/67d2c590926572660a4d2b6c)

### [Gemma 3: Advancing Open-Source Multimodal AI with Scalable Training and Efficient Architectures](https://airesearch.hashnode.dev/gemma-3-advancing-open-source-multimodal-ai-with-scalable-training-and-efficient-architectures?utm_source=hashnode&utm_medium=feed)

Discover the groundbreaking advancements of Google's Gemma 3 in open-source multimodal AI. Explore its scalable training, efficient architectures etc.

15 min read · [airesearch.hashnode.dev](https://airesearch.hashnode.dev/gemma-3-advancing-open-source-multimodal-ai-with-scalable-training-and-efficient-architectures?utm_source=hashnode&utm_medium=feed)
