How to Protect Your Privacy From a Generative AI Chatbot Company

Generative AI tools like ChatGPT have been heralded as a new way of doing everything, from streamlining our lives to helping businesses function better. Generative AI in the form of a website chat tool or an automated assistant has been compared to the Industrial Revolution, but it's important to remember that with any new tool that promises to change the world, we must learn how to use it properly.

One of the biggest concerns with tools like ChatGPT is privacy. There are a number of risks with generative AI tools like ChatGPT, including:

• Your data may have been used to train ChatGPT, as it was fed 300 billion words with the aim of generating human-like responses, collating all types of content online from comments, blog posts, books, texts, and articles.
• Your conversations are recorded and used to train ChatGPT, resulting in a potential leak of personal information.
• ChatGPT can share information with third parties such as vendors, business transfers, affiliates, and law enforcement agencies.

Generative AI raises a lot of concerns about data privacy because it can process personal data and generate potentially sensitive information, and it could inadvertently be collecting personal information, and this is where we should all proceed with caution. ChatGPT was briefly banned in Italy due to privacy concerns, but the rest of the world is able to use it to their advantage. So, what can we do to protect our privacy while using generative AI tools? There are many ways to do this:

Be Careful With Personal Data

Using platforms like ChatGPT and other generative AI means that we've got to think twice about any sensitive information, whether it's personal information like names, addresses, or photos. We must consider the bigger picture of how this information can be used. While many businesses are using generative AI as a website chat tool or chatbot software in general, it is critical for users to be more considerate of what they should share.

When it comes to sharing sensitive information on business platforms, there are ways around this, which we will get into, but as a general rule, generative AI can inadvertently aid phishing, and these attempts have skyrocketed over the last few months. Users should think about what they want to share while also protecting online accounts with tools such as two-factor authentication or biometrics.

Use a VPN (Virtual Private Network)

This is pivotal for customers or users that want to protect any sensitive information. A VPN can prevent AI from tracking a user across the internet by providing an encrypted connection and anonymizing the IP address.

When a user or website provider connects to a VPN server, the IP address is hidden, and as users are assigned a temporary IP address, this will alter the true identity and location, which makes it harder for websites to track online activities as well as generative AI use.

Be Careful in the Office!

As useful as generative AI can be in the office environment, there have been instances of ChatGPT conversations being leaked. For example, Samsung's Semiconductor engineering team pasted confidential code onto ChatGPT twice, which was promptly absorbed by ChatGPT as training data! Samsung learned the hard way and is now creating its own AI systems, but while this is not possible for everybody, we should maintain some sense of healthy practice when interacting with tools like ChatGPT. For example, use generic identifiers rather than personal information to increase anonymity and not associate generated content with the real person behind it.

Employees must also think about reporting any concerns, and if there are any privacy concerns or a breach of information while using a generative AI tool, it is pivotal to report the issues either to the service provider or to go above them and to the regulatory authority.

Prioritise Secure Connections and Encryption

Users can benefit from a VPN, but the generative AI service must also have a duty of care and ensure that it applies encryption to protect state transmission between the device and server. One of the best ways to do this is to check if the URL has "HTTPS" as part of its address, which highlights that it is encrypted.

It's also important to look at the track record of the AI tool in handling user data. ChatGPT has recently experienced a massive leak of credentials, with over 100,000 accounts affected and sold on dark web marketplaces between June 2022 and May 2023. Therefore, many users of generative AI tools like ChatGPT are thinking twice. As useful as they can be, any sensitive information on these tools can suffer from misuse.

Review Privacy Policies

Generative AI tools will all provide privacy policies, and it's important to understand how your information will be collected, stored, and used by the system. ChatGPT states that they may use the data to improve their models, but OpenAI offers an opt-out form. Many users bypass this; however, if you're looking for information on how your data is collected, stored, retained, and shared, you might be surprised. Prioritise privacy clarity and always be sure to have security measures in place.

You can also contact OpenAI directly to request that ChatGPT remove personal information, for example, via their dedicated removal form or by sending an email to dsar@openai.com, but you can also avoid sharing too much information when interacting with the chatbot company. While any business that uses a website chat tool like a chatbot should have all the best practices for privacy in place, there is the potential for any misuse of information.

Generative AI tools like ChatGPT have undoubtedly changed how we all operate, and since launching in November 2022, ChatGPT became the fastest-growing platform of all time, but despite its clear advantages, there is still the privacy debate that rages on, which is why it's more important than ever to become aware of how to protect your information but also be cautious with any new generative AI tool. By understanding the technology, being cautious with data, understanding and reviewing privacy policies, and, of course, being vigilant, anybody can safeguard their privacy with generative AI.