What is Generative AI?
Artificial Intelligence (AI) refers to the capabilities of machines, particularly computers, to mimic human problem-solving, decision-making, and task performance. Generative AI is a type of artificial intelligence technology that can produce new content such as images, text, audio, programming code, and artworks.
Watch: What is AI? (4:38) Watch: Generative AI Explained (2:02)
AI, Privacy, and You
Generative AI tools use the information that users input into them (including your prompts) to continue to learn and fine-tune their performance. A ‘prompt’ is any form of text, information, coding, or question that conveys to the AI tool what response or content you are looking for. In this sense, your data becomes the AI’s data. Once you enter information into one of these models, there is no guarantee that it will not become publicly available in unpredictable ways.
Like most technologies, there are both benefits and risks attached to the use of Generative AI. An important consideration when using AI tools is your data privacy and the risk of malicious cyberattacks and data breaches. ChatGPT has already been subject to a data breach which included a leak of personal information such as individuals’ names, addresses, private chatbot messages, and payment information. See the full article.
To better protect your privacy and sensitive data in relation to using AI tools, check out the tips and recommendations below:
Privacy Tips and Recommendations for Safer Use of Generative AI:
- Only use a trusted AI tool and access it only through its official channel.
Don’t download unverified AI links, applications, or media, especially those shared on public hubs and forums such as Discord, Reddit, or GitHub.
- When creating your account(s) with generative AI tools don’t sync it with any pre-existing account you may have such as Google, Microsoft, Apple, of Facebook accounts.
- Limit the personal information you share in these models, such as date of birth, full name, address, and phone number. Never share information about your finances, health status, social security number, or other sensitive data with AI tools.
- Do not share or input voice recordings or identifiable images of yourself into AI tools.
- Understand the data controls and settings of the tool you are using.
Some tools, like ChatGPT, allow you to modify settings to delete your chat history or restrict if the model can use your inputs/prompts to improve its performance.
- Read Privacy Policies. Understanding how companies protect (or don’t protect!) your privacy is essential to making smart, safe decisions to prevent your data being breached and to protect yourself from identity theft.
- Be cautious about AI-generated messages and report inappropriate content.
If you receive unexpected or suspicious messages from AI chatbots or accounts, do not engage in conversation or provide any information requested. Report all unusual activity to the platform or service provider.
ChatGPT: What to be Aware of Regarding Privacy
OpenAI’s ChatGPT is a widely popular generative AI Tool. Within five days of its release in November 2022, it reached millions of users. It has since had more recent and powerful versions released (ChatGPT 4.0). Both free and paid versions of ChatGPT are available to users.
ChatGPT is a natural language processing tool that operates through an interactive chatbot interface. It can assist with tasks like writing emails, code, essays, songs, etc. It can also assist with research, synthesize data, and run large search queries. See below for important privacy and data protection information as found within OpenAI’s own Privacy Policy:
OpenAI (ChatGPT & Dall-e)
Data Collection, Use, and Sharing
- OpenAI collects personal information you provide. This includes your name, contact info, payment info, files and content you upload, prompts, messages you send to OpenAI, and activities on social media accounts you may visit.
- OpenAI may use User Content you provide to improve its services. User content is the input, file uploads, or feedback that you provide to OpenAI and may include personal information if you have supplied it. User content may be used for the purposes of training the models that power ChatGPT and/or Dall-e.
You can switch off training in ChatGPT settings (under Data Controls) to turn off training of any conversations created. This will limit OpenAI from using your data to improve its model performance
You can also delete chat conversation histories under Account Settings.
- OpenAI collects personal information received automatically from your use of their services. This includes log data (your IP address, date, etc.), usage data (types of content you viewed, features you engaged with, date/time/country of access), device information (name of your device, operating system, device identifiers), and tracking cookies.
- OpenAI’s Mobile App may collect personal information and data. This includes geolocation, name, email address, in-app messages, app activity, and user-generated content.
- OpenAI states that they "may provide your Personal Information to third parties without further notice to you". These third parties include vendors, other providers, law enforcement, and affiliates.
OpenAI does not guarantee the privacy of your personal information and may share it for commercial or legal purposes. Although OpenAI claims they do not intend to use personal information, they have acknowledged that personal information supplied to them may be used to further train their AI models. This likely applies to other AI tools in a similar manner.
For more information about OpenAI and ChatGPT see their Privacy Policy, Terms of Use, and statement on their approach to AI safety.
For more information as to your privacy rights in B.C. in relation to private organizations, like OpenAI, check out this Guide to the Personal Information Protection Act (PIPA) or see here.
If you have any questions or concerns about privacy and personal information at Douglas College please see our website, or contact the Privacy Office at privacy@douglascollege.ca.