Hot Off the Presses: The Meta AI chatbot trained with user posts, and ChatGPT is no longer limited to September 2021 data. Follow along with Sharon Nelson, John Simek and Michael Maschke as they explain what that means for users.
Table of contents
- Hurray! ChatGPT Now Has Current Data
- How Do You Get ChatGPT to Browse Bing?
- Meta Confesses That the Meta AI Chatbot Was Trained on Public Facebook and Instagram Posts — Both Texts and Photos
- Securing AI
- More on ChatGPT for Lawyers
Lawyers complained that ChatGPT was not entirely useful to them because it was limited to data prior to September 2021 — a huge limitation for an attorney doing research or anything requiring access to data after 2021. It really prevented a lot of lawyers from using ChatGPT in any major way.
Good news: Late last month, OpenAI announced that ChatGPT can now browse the internet.
Hurray! ChatGPT Now Has Current Data
We have been stymied time and again by not having current data from ChatGPT. But this week we were able to ask ChatGPT how Congress came to pass a short-term budget bill to avoid a government shutdown, and it had all the most current information.
We could even determine what reliable sources are saying about the possibility of passing a new budget: “The likelihood of Congress passing a new budget before the short-term budget bill expires seems uncertain.” And it told us why. Importantly, we could also see which sources it was consulting and that they were all well-respected ones.
To further test-drive it, we asked for the most significant legal news on October 2, 2023, at noon. ChatGPT chose the lawsuit against Amazon by the Federal Trade Commission and 17 states, which alleges Amazon abuses its market position to inflate prices, overcharge sellers and stifle competition. Significant news indeed.
It was hard to stop playing with ChatGPT to pen this article!
How Do You Get ChatGPT to Browse Bing?
For author Nelson, that was the hardest part to figure out — which is where author Simek proved very useful. Here’s how you do it — and remember that, for the moment, the update is available only to premium users on ChatGPT’s “Plus” and “Enterprise” plans. However, OpenAI says the update will “expand to all users soon,” but it did not give the timing of the update’s rollout.
First, you have to “turn on” the ability for ChatGPT to access the internet by configuring it to browse using Bing.
- From the main screen, click on the GPT-4 button at the top and click on the “Browse with Bing” option.
- A checkmark will display next to the option to let you know ChatGPT will now use Bing to get to the internet for current information.
Why did Bing get picked as the browser?
Well, Microsoft owns 49% of OpenAI. Is the picture clearer now?
Add this to the fact that ChatGPT can now “speak,” listen and process images and you have a well-rounded tool at your command.
Meta Confesses That the Meta AI Chatbot Was Trained on Public Facebook and Instagram Posts — Both Texts and Photos
This September 2023 revelation did not sit well with a lot of people, including lawyers. Meta AI is still in beta and Meta said it excluded private posts shared with family and friends to preserve consumers’ privacy.
Meta President of Global Affairs Nice Clegg said, “We’ve tried to exclude datasets that have a heavy preponderance of personal information.”
Meta, Open AI and Alphabet’s Google have all been severely criticized for scraping data from the internet without permission to train their AI models.
As readers are likely aware, many book authors have filed lawsuits against these companies alleging that they are guilty of copyright infringement. It will be very interesting to see the outcomes of these lawsuits.
A lot of lawyers have asked about securing AI. From the lawyer’s point of view, you should always designate your conversations with AI as private and not to be used for training.
Disabling Chat History in ChatGPT
For ChatGPT, which is what we use, you disable your chat history and opt out of having your ChatGPT data used to improve the models — you toggle the Data Controls setting “Chat History & Training” to OFF, whether you use the free or paid version of ChatGPT.
There Is Some AI Security Lawyers Cannot Control
For instance, AI companies struggle with such things as training data poisoning. Malicious actors can feed the AI tool flawed data or corrupt legitimate training data. This is obviously beyond the lawyer’s control, but lawyers must be aware of the possibility that the AI output may be corrupted.
AI companies are aware of the dangers and, presumably, acting to monitor for data poisoning and other AI security hazards.
There are many guides to implementing a secure AI framework that are entirely beyond the understanding of most lawyers. This is why we recommend that lawyers use ONLY AI that has established a reputation for security. (See our Attorney at Work article “Beware of Ethical Perils When Using Generative AI.”)
As law firms begin rapidly using AI, training law firm employees on the use of AI is becoming a critical component of law firm security — and training is too often omitted. While that’s another column entirely, be mindful of the dangers of failing to train.
When working with AI, it is often asked, “What could possibly go wrong?” The answer is a lot.
Sharon D. Nelson is a practicing attorney and the president of Sensei Enterprises, Inc. She is a past president of the Virginia State Bar, the Fairfax Bar Association and the Fairfax Law Foundation. She is a co-author of 18 books published by the ABA. firstname.lastname@example.org.
John W. Simek is vice president of Sensei Enterprises. He is a Certified Information Systems Security Professional (CISSP), Certified Ethical Hacker (CEH) and a nationally known expert in digital forensics. He and Sharon provide legal technology, cybersecurity and digital forensics services from their Fairfax, Virginia, firm. email@example.com.
Michael C. Maschke is the CEO/Director of Cybersecurity and Digital Forensics of Sensei Enterprises. He is an EnCase Certified Examiner and a Certified Computer Examiner. firstname.lastname@example.org.
More on ChatGPT for Lawyers
- “Harnessing ChatGPT: A Primer for Lawyers” by Mark C. Palmer
- “Is Your Firm Generative AI Ready?” by Alex Smith
- “Make AI Your Intern, Not Your Replacement as a Lawyer” by Ruth Carter
- “Beware of Ethical Perils When Using Generative AI” by Sharon Nelson, John Simek & Michael Maschke
- “Generative AI for Law Firm Marketing” by David Arato
Image © iStockPhoto.com
Don’t miss out on our daily practice management tips. Subscribe to Attorney at Work’s free newsletter here >
More Cybersecurity Tips:
- Ransomware Today: Top Tips for Law Firms (Get these 14 things done and you’re way ahead of most of your colleagues)
- Cybersecurity Trends: 25% of Law Firms Have Been Breached
- Shadow IT: A Serious Threat to Law Firms
Subscribe to Attorney at Work
Get really good ideas every day for your law practice: Subscribe to the Daily Dispatch (it’s free). Follow us on Twitter @attnyatwork.