Safety.AI
A Global Standard for Artificial Intelligence
カートのアイテムが多すぎます
カートに追加できませんでした。
ウィッシュリストに追加できませんでした。
ほしい物リストの削除に失敗しました。
ポッドキャストのフォローに失敗しました
ポッドキャストのフォロー解除に失敗しました
プレミアムプラン4か月 月額99円キャンペーン開催中
¥2,750 で購入
-
ナレーター:
-
MR Daniel Buxton
-
著者:
-
Daniel Giedroyce
Safety.AI
Building the Global Standard for Safe, Secure, and Responsible Artificial Intelligence
AI capabilities are advancing faster than the standards meant to govern them. Today's systems make hiring decisions, diagnose diseases, write software, drive vehicles, detect fraud, generate synthetic media, and act autonomously — often without the safety engineering, governance, and certification expected in aviation, medicine, or cybersecurity.
Safety.
AI argues that AI needs its own universally recognized safety standard. Drawing on lessons from cybersecurity, software engineering, critical infrastructure, and real-world AI failures, this book introduces the Safety.
AI Framework — a comprehensive approach to designing, deploying, governing, auditing, and monitoring AI systems throughout their lifecycle.
Inside you'll learn:
• Why AI needs a different safety paradigm than traditional software
• AI's evolution from machine learning to LLMs, generative AI, and autonomous agents
• How bias, hallucinations, deepfakes, prompt injection, model poisoning, and alignment failures create new organizational risks
• Lessons from real-world incidents involving Amazon, Google, Tesla, Uber, ChatGPT, Clearview AI, Cambridge Analytica, and others
• How NIST AI RMF, ISO/IEC 42001, UL 3115, and emerging regulations fit the governance landscape
• Practical guidance on risk, security, privacy, fairness, explainability, and human oversight
• A proposed Safety Level classification system for certifying AI by risk and impact
Safety.
AI offers a practical roadmap for organizations deploying AI across healthcare, finance, government, manufacturing, transportation, education, and other critical industries. Whether you're an executive, engineer, compliance officer, or policymaker, this book gives you the tools to build AI that's powerful and trustworthy.
As AI becomes critical infrastructure, safety can no longer be optional.
PLEASE NOTE: When you purchase this title, the accompanying PDF will be available in your Audible Library along with the audio.
©2026 Daniel Giedroyce (P)2026 Daniel Giedroyce