際際滷

際際滷Share a Scribd company logo
Linkedin.com/in/mohamed-helmy
Linkedin.com/in/mohamed-helmy
What you MUST know
about Privacy on ChatGPT
Linkedin.com/in/mohamed-helmy
Linkedin.com/in/mohamed-helmy
Conversation may be
read by AI trainers
In my recent survey, 57% of people didn't know
that real humans - aka AI trainers - can read their
conversations, although it's one of the hints you
get when you 鍖rst sign up.
As data privacy concerns continue to rise, a new
discussion has emerged regarding the risks of
sharing con鍖dential information with AI-based
chatbots like ChatGPT. While the technology has
made strides in various 鍖elds, its potential
drawbacks can't be ignored.
Linkedin.com/in/mohamed-helmy
Linkedin.com/in/mohamed-helmy
Con鍖dential information
is being shared through
ChatGPT conversations
In a recent example involving Samsung, con鍖dential
information was unintentionally shared through
ChatGPT, raising concerns about the potential risks of
using AI-based platforms. Samsung's response was to
limit employee access to the AI, a move that signals the
increasing need for organizations to prioritize data
privacy. Another privacy concerns were raised in Italy
by the countrys data protection watchdog, followed by
a ban of the tool in Italy.
Linkedin.com/in/mohamed-helmy
Linkedin.com/in/mohamed-helmy
Who owns your input and
AIs output
As highlighted by Senior Council Samantha Schonfeld on
LinkedIn, there are legal implications associated with
AI-generated content. As per the terms; the user owns
any Input provided to ChatGPT as well as any Output
that ChatGPT provides in response. However, OpenAI
may use Content [de鍖ned as Input and Output] as
necessary to provide and maintain the Services.
We need to ask: Who is responsible for mistakes made by
AI? Can these errors lead to copyright infringement or
other legal liabilities?
Linkedin.com/in/mohamed-helmy
Linkedin.com/in/mohamed-helmy
Action Plan
1
Blocking ChatGPT (Plugins) with Robots.txt
To implement this, simply add the following
lines to your robots.txt 鍖le:
User-agent: ChatGPT
Disallow: /
User-agent: ChatGPT-User
Disallow: /
This will prevent the AI plugins from crawling
and accessing your website's content, until
there's a clear revenue-sharing model
between OpenAI and Content providers.
2
Stop ChatGPT from Reading Conversations
Request to opt-out of having your data used
by 鍖lling out this form with your organization
ID and email address associated with the
owner of the account. This applies to both
individuals and companies. You can also;
1. Delete any sensitive or con鍖dential
information from your conversation
history.
2. Avoid sharing sensitive information in
future conversations.
Linkedin.com/in/mohamed-helmy
Linkedin.com/in/mohamed-helmy
Stay Safe
1
Educate your team
Ensure that everyone in your organization
understands the risks and best practices
associated with AI platforms like ChatGPT.
2
Establish clear guidelines
Develop a set of rules for using ChatGPT. This
might include outlining acceptable use cases,
specifying who has access, and providing
instructions for reporting any issues.
3
Monitor usage
Regularly review how ChatGPT is being used
within your organization to ensure compliance
with established guidelines and to identify any
potential risks.
Linkedin.com/in/mohamed-helmy
Linkedin.com/in/mohamed-helmy
Thank you.
Linkedin.com/in/mohamed-helmy
bit.ly/MarketersNewsletter

More Related Content

What you MUST know about Privacy on ChatGPT - Mohamed Helmy.pdf

  • 2. Linkedin.com/in/mohamed-helmy Linkedin.com/in/mohamed-helmy Conversation may be read by AI trainers In my recent survey, 57% of people didn't know that real humans - aka AI trainers - can read their conversations, although it's one of the hints you get when you 鍖rst sign up. As data privacy concerns continue to rise, a new discussion has emerged regarding the risks of sharing con鍖dential information with AI-based chatbots like ChatGPT. While the technology has made strides in various 鍖elds, its potential drawbacks can't be ignored.
  • 3. Linkedin.com/in/mohamed-helmy Linkedin.com/in/mohamed-helmy Con鍖dential information is being shared through ChatGPT conversations In a recent example involving Samsung, con鍖dential information was unintentionally shared through ChatGPT, raising concerns about the potential risks of using AI-based platforms. Samsung's response was to limit employee access to the AI, a move that signals the increasing need for organizations to prioritize data privacy. Another privacy concerns were raised in Italy by the countrys data protection watchdog, followed by a ban of the tool in Italy.
  • 4. Linkedin.com/in/mohamed-helmy Linkedin.com/in/mohamed-helmy Who owns your input and AIs output As highlighted by Senior Council Samantha Schonfeld on LinkedIn, there are legal implications associated with AI-generated content. As per the terms; the user owns any Input provided to ChatGPT as well as any Output that ChatGPT provides in response. However, OpenAI may use Content [de鍖ned as Input and Output] as necessary to provide and maintain the Services. We need to ask: Who is responsible for mistakes made by AI? Can these errors lead to copyright infringement or other legal liabilities?
  • 5. Linkedin.com/in/mohamed-helmy Linkedin.com/in/mohamed-helmy Action Plan 1 Blocking ChatGPT (Plugins) with Robots.txt To implement this, simply add the following lines to your robots.txt 鍖le: User-agent: ChatGPT Disallow: / User-agent: ChatGPT-User Disallow: / This will prevent the AI plugins from crawling and accessing your website's content, until there's a clear revenue-sharing model between OpenAI and Content providers. 2 Stop ChatGPT from Reading Conversations Request to opt-out of having your data used by 鍖lling out this form with your organization ID and email address associated with the owner of the account. This applies to both individuals and companies. You can also; 1. Delete any sensitive or con鍖dential information from your conversation history. 2. Avoid sharing sensitive information in future conversations.
  • 6. Linkedin.com/in/mohamed-helmy Linkedin.com/in/mohamed-helmy Stay Safe 1 Educate your team Ensure that everyone in your organization understands the risks and best practices associated with AI platforms like ChatGPT. 2 Establish clear guidelines Develop a set of rules for using ChatGPT. This might include outlining acceptable use cases, specifying who has access, and providing instructions for reporting any issues. 3 Monitor usage Regularly review how ChatGPT is being used within your organization to ensure compliance with established guidelines and to identify any potential risks.