*Last updated: July 2024*
<aside>
<img src="/icons/wheat_green.svg" alt="/icons/wheat_green.svg" width="40px" /> The Mindful AI circle was established with the aim of delving into the mindful utilisation of AI on both personal and organisational levels within ATÖLYE.
We believe that AI has the potential to revolutionise the way we live and work, but we also recognise that it's still developing and that it's not yet fully trustworthy. As such, we seek to promote a more mindful approach to AI by exploring its use in a comprehensive and ethical manner to ensure that this powerful technology is used in a way that benefits individuals, organisations, and society as a whole.
We hope that this guide will help you leverage the power of human-AI collaboration rather than being dependent on or afraid of generative AI tools.
</aside>
<aside>
<img src="/icons/bullseye_yellow.svg" alt="/icons/bullseye_yellow.svg" width="40px" /> Our primary objective is to publish a comprehensive guide that examines aspects of privacy, security, confidentiality, and ethics, accompanied by a set of principles, along with "Dos" and "Don’ts".
Additionally, we will present **specific** **use cases** to complement the information outlined in this guide.
</aside>
<aside>
<img src="/icons/soccer_orange.svg" alt="/icons/soccer_orange.svg" width="40px" /> This is a living page that continues to develop. Some content such as resource database, checklist to assess the reliability of an AI tool, ATÖLYE Use Cases, etc. ****will be improved.
</aside>
What Is Artificial Intelligence and How Does It Work?
Recommended Generative AI Usage
Fact-check.
Paraphrase.
Give proper attribution when necessary (link, citation, etc.).
Use pseudonyms for any personal info.
Educate yourself about the latest developments, ethical guidelines, and best practices in AI usage.
Be transparent, cite the source.
Share sensitive information.
Share harmful or misleading content.
Rely on a single AI tool.
Use AI-generated content without verifying it.
Disregard bias.
Neglect intellectual property rights.
Do not use ATÖLYE or Client IP and Confidential Information.
<aside>
1️⃣ Confidentiality: Do not share sensitive company information or confidential data such as confidential project details, financial information, employee data, client information, strategic plans, etc. without anonymising the data.
<aside> <img src="/icons/info-alternate_yellow.svg" alt="/icons/info-alternate_yellow.svg" width="40px" /> AI tools, including chatbots, virtual assistants, and other generative AI models, are designed to process the data they receive to perform tasks. The process usually involves analysing patterns in the data and making associations based on those patterns. Therefore, when sensitive or confidential data is input into these tools, it's crucial to ensure that the AI tool in question is specifically designed and approved to handle such information securely and confidentially.
</aside>
<aside> <img src="/icons/magic-wand_blue.svg" alt="/icons/magic-wand_blue.svg" width="40px" /> While using ChatGPT with your personal account, you can opt out of model training by turning off the chat history & training option. Please keep in mind that this means you won’t be able to access your conversation history for new chats.
</aside>
<aside> <img src="/icons/emoji-surprised_gray.svg" alt="/icons/emoji-surprised_gray.svg" width="40px" /> The Economist Korea reported three separate instances of Samsung employees unintentionally leaking sensitive information to ChatGPT. […] A third, shared a recording of a meeting to convert into notes for a presentation. That information is now out in the wild for ChatGPT to feed on.
Whoops, Samsung workers accidentally leaked trade secrets via ChatGPT
</aside>
</aside>
<aside>
2️⃣ Privacy protection: Do not share personal information of clients, colleagues, or any other individual with AI tools.
<aside> <img src="/icons/info-alternate_yellow.svg" alt="/icons/info-alternate_yellow.svg" width="40px" /> Personal information includes, but is not limited to, names, addresses, contact details, health information, financial data, and any other information that could be used to identify or locate a person.
Even if an AI tool doesn’t intentionally store this data, simply processing it may expose personal information to potential risks like data breaches, data misuse, identity theft, invasion of privacy, etc.
</aside>
</aside>
<aside>
3️⃣ Anonymisation: Whenever possible, anonymise data by erasing or encrypting, before using it with AI tools to minimize the risk of exposing personally identifiable information (PII) or other sensitive data. This principle is especially crucial when using AI tools in the workplace as it significantly reduces the risk of exposing sensitive data.
</aside>
<aside>
4️⃣ Verification: When using AI-generated content, ensure it comes from a reputable source or tool to minimize the risk of misinformation or manipulation. If you asked for citations, references, or URLs, or if the content includes factual claims, check their validity and accuracy by verifying them with other sources.
<aside> <img src="/icons/warning_yellow.svg" alt="/icons/warning_yellow.svg" width="40px" /> The current ChatGPT model isn’t capable of giving insights after September 2021. Please consider that while using content generated by ChatGPT.
However, you can use GPT-4’s web browsing extension to access real-time information.
</aside>
<aside> <img src="/icons/search_green.svg" alt="/icons/search_green.svg" width="40px" /> Try it yourself! Here’s a recent ChatGPT conversation 4️⃣0️⃣4️⃣
Certainly, here are some additional resources that discuss AI ethics, safety, and best practices. These resources can help inform the creation of rules for using AI tools in the workplace:
These resources should provide a broader understanding of AI ethics, safety, and best practices, which can help you create guidelines for using AI tools in your organization.
</aside>
</aside>