AI Best Practice Resources

A list curated by the Greater Manchester AI Foundry¹ to provide resources on and promote best practice with respect to the development of AI.

Links have been selected with the aim of helping AI developers navigate the latest regulations, practices, and research. However, this list will never be comeplete and we may have missed things!

The contribution guidelines can be found here.

Governing / Advisory Bodies

• 🌐 The Alan Turing Institute
• 🌐 The Institute for Ethical AI & Machine Learning
• 🌐 Centre for Data Ethics and Innovation
• 🌐 European AI Alliance
• 🌐 Ada Lovelace Institute

Regulation / Strategy

• 🌐 National AI Strategy (UK national AI strategy published 2021)
• 🌐 Algorithmic Transparency Standard
• 📜 Auditing algorithms: the existing landscape, role of regulators and future outlook
• 📜 AI, human rights, democracy and the rule of law: A primer prepared for the Council of Europe
• 📜 European Commission - Ethics and Data Protection

Toolkits / Best Practice

• 🛠️ Microsoft Responsible Innovation Toolkit
• 🌐 The Time Is Now to Develop Community Norms for the Release of Foundation Models
• 📜 Co-Designing Checklists to Understand Organizational Challenges and Opportunities around Fairness in AI
• 📜 Building Trustworthy AI Solutions: A Case for Practical Solutions for Small Businesses
• 📜 Microsoft Azure AI Security Risk Assessment
• 📜 Datasheets for Datasets
• 🎬 Lies, Distortions, and Misrepresentations in Visualisation (Michael Correll discusses bad practice in data visualisation that can mislead and audience)
• 🎬 Use of Colour (Cole Knafflic talks about the use of colour in good data vis design)
• 🎬 Use of Contrast (Cole Knafflic talks about the use of contrast in good data vis design)
• 🎬 How to Declutter (Cole Knafflic talks about how to declutter in good data vis design)

Legal, Ethical, & Social Implications

• 📜 The Top 10 Risks of Machine Learning Security
• 📜 On the Dangers of Stochastic Parrots
• 📜 Interpretable machine learning: Fundamental principles and 10 grand challenges

Footnotes

1. This work is a direct result of working on the ERDF Greater Manchester AI Foundry which is part funded by the European Regional Development Fund. ↩