The Gist

  • Changing landscape. OpenAI’s ChatGPT changed the landscape for generative AI applications.
  • Musk threatens lawsuit. Elon Musk has threatened to sue OpenAI for using Twitter data to train ChatGPT.
  • Future in Europe? Italy, the first European country to ban ChatGPT, has reinstated its use after OpenAI addressed watchdog's concerns.

Generative AI has taken the world by storm since the release of OpenAI’s ChatGPT in November of 2022. Since then, the AI organization has released the latest version, GPT4, which has additional functionality, and many businesses have rushed to create applications with GPT functionality for a multitude of industries. Because of such widespread interest in the new AI technology and the strength of the large language model, many people have expressed concern over the potential of AI to impact the future of humanity, something that until recently could have been considered to be part of a dystopian science-fiction novel. 

Elon Musk Is Feeling the FOMO

Twitter CEO Elon Musk rarely holds back when he feels strongly about something and recently tweeted that OpenAI’s use of Twitter for training ChatGPT was going to result in a lawsuit.


Musk, who co-founded OpenAI in 2015 and was one of the original board members until he stepped down from the company's board in 2018, recently signed an open letter (i.e., a petition) to halt AI development for six months. Not long after publishing the letter, which as of April 2023 had 27,567 signatures, Musk announced that he was going to create his own generative AI chatbot called TruthGPT, a nod to former President Trump’s Truth Social network, and an indication that Musk believes that ChatGPT is liberally biased. On the now canceled Fox News show,“Tucker Carlson Tonight,” Musk related that he’s going to “start something which I call TruthGPT,” stating that he’d want his generative AI app to be a “maximum truth-seeking AI that tries to understand the nature of the universe.” 

The fact that OpenAI used Twitter as part of a multitude of websites that were crawled in order to train the generative AI model is not in dispute. However, the percentage that it contributed is extremely small (0.0008%), as can be seen in this tool that the Washington Post provided, which enables users to search for a website in the crawl data:

washington post

In natural language processing (NLP) and AI, tokens are the smallest units of meaning in a text, and can be words, subwords, phrases or characters. Tokens serve as the input for AI models that process text.

Related Article: ChatGPT Suffers First Data Breach, Exposes Personal Information

ChatGPT and Large Language Models

Large language models are not entirely new, but OpenAI’s ChatGPT changed the landscape for generative AI applications. Because of the interest that was generated by the announcement, many businesses have rushed to bring their own generative AI applications to market without considering the overall implications that come with the technology.  

All large language models, as their name would suggest, are trained on very large amounts of data, most of which is scraped from the internet. As such, much of that data is owned (and copyrighted) by other businesses and individuals. This leaves the question, as Musk tweeted, as to whether the data was illegally used to train the large language model.

In addition, Microsoft, GitHub and OpenAI are being sued as part of a class action lawsuit that accuses them of violating copyright law through the use of Copilot, a code-generating AI system that was trained on billions of lines of public code, which enables it to regurgitate licensed code snippets without providing credit. 

OpenAI is also being served by Gordon Legal, as reported by Ars Technica, for a possible defamation claim based on ChatGPT 3.5 generating "several false statements" stating that Brian Hood, a regional mayor in Australia, "was accused of bribing officials in Malaysia, Indonesia, and Vietnam between 1999 and 2005, that he was sentenced to 30 months in prison after pleading guilty to two counts of false accounting under the Corporations Act in 2012, and that he authorized payments to a Malaysian arms dealer acting as a middleman to secure a contract with the Malaysian Government." However, Hood's legal teams says that Hood was instead the person who notified authorities about the bribes.

Learning Opportunities

Related Article: ChatGPT Is All the Rage but Don't Stop Learning Just Yet

Italy First European Country to Ban ChatGPT

On March 31, 2023, the Italian Data Protection Authority Garante ordered OpenAI to stop processing Italian users’ data amid a data breach that enabled users to see the titles of conversations other users were having with the chatbot. The ban was largely put in place because ChatGPT did not comply with the European General Data Protection Regulation (GDPR). As such, Italy was the first European country to ban ChatGPT.

Just before the April 30 deadline it imposed, Garante lifted the ChatGPT ban after OpenAI introduced changes to comply with data privacy requirements. OpenAI enhanced its website to provide clearer information on how the chatbot processes user data and introduced a feature for users to opt-out of having their conversations used in training ChatGPT's algorithms.

Other countries to ban the AI chatbot include:

  • China
  • Cuba
  • Iran
  • North Korea
  • Russia
  • Syria

Some of these countries’ respective governments banned AI chatbots based on privacy concerns, and others, such as North Korea, China and Russia, ironically felt that the US would use ChatGPT to spread misinformation. Additionally, OpenAI does not allow the following countries to use ChatGPT:

  • Afghanistan
  • Bhutan
  • Central African Republic
  • Chad
  • Eritrea
  • Eswatini
  • Iran
  • Libya
  • South Sudan
  • Sudan
  • Syria
  • Yemen

Final Thoughts on the Future of AI

Concerns about ethics, copyright infringement, privacy and defamation are not new, and like any other transformative technology, there are going to be challenges that will need to be solved as the technology continues to evolve. The proverbial genie is out of the bottle, however, and it’s highly unlikely that any of these challenges will mean the demise of generative AI and large language models.