Question 1

How can I learn to build responsible and secure AI tools and services?

Accepted Answer

Microsoft Foundry (officially Azure AI Foundry) is a unified Azure platform-as-a-service for enterprise AI development, operations, and application building — combining infrastructure, MLOps, model access, and tools in one place. (Microsoft Learn)(https://learn.microsoft.com/en-us/azure/ai-foundry/what-is-azure-ai-foundry?view=foundry-classic&utm_source=chatgpt.com)
Amazon Bedrock + SageMaker provide a managed platform for generative AI and foundation models (LLMs) that lets you access and use a variety of powerful pre-trained models (e.g., Claude, Llama, etc.) through a unified API. (AWS Documentation)(https://docs.aws.amazon.com/bedrock/?utm_source=chatgpt.com) and an end-to-end machine learning platform for model training, tuning, deployment, and monitoring. It complements Bedrock by giving developers full machine learning (ML) lifecycle control
Google Vertex AI is Google Cloud’s unified platform for building, training, deploying, and managing machine learning and AI (including generative AI and foundation models). It includes tools like Vertex AI Studio, Model Garden, and tutorials for working with models, MLOps workflows, and AI pipelines. (Google Cloud Documentation)(https://docs.cloud.google.com/vertex-ai/docs?utm_source=chatgpt.com). Vertex AI plays the same overall role on GCP that Foundry does on Azure — a one-stop platform for enterprise AI and ML development.

Question 2

Is there any government guidance I could follow?

Accepted Answer

The UK Government National Cyber Security Centre (NCSC) has developed guidelines for secure AI system development (https://www.ncsc.gov.uk/collection/guidelines-secure-ai-system-development)

The United States National Institute of Standards and Technology (NIST) has developed a comprehensive  AI risk management framework t(https://airc.nist.gov/)hat includes a risk management framework, profiles and use cases.