One of the newest innovations in technology, ChatGPT is a large language model (LLM) with the ability to generate human-like responses to a wide range of questions and prompts. It is the fastest growing application ever, amassing over 100 million users in 2 months. Undoubtedly this is an incredible innovation, however there are some concerning risks associated with it that affect not only large organisations but individuals too.
Privacy
A primary concern with ChatGPT is the issue of privacy and protection of personal data. Since it is trained on a vast amount of training data initially, there is a very high probability that the AI chatbot can access sensitive information. There needs to be some sort of solid consent model to protect users.
Currently there seems to be an ‘open slather’ style model whereby a user has to essentially sign their life away (not literally) in order to use the product. Going forward this is a standard I believe won’t be widely accepted. Social policy and legal systems are put in place to protect user rights, even if it ‘lags’ the development of the technology and it is about time they are used.
In addition to this every piece of incoming data from new users is stored by the parent company ‘OpenAI’. If this information was acquired by someone with malicious intent, it would pose a significant threat such as exploitation against the very individuals and organisations it originated from.
Additionally, due to the AI’s ability to mimic human like responses, scams, phishing, and other attacks could become increasingly automated and highly sophisticated. A concerning example is an AI powered piece of malware. Generative AI has been maliciously used by hackers and fraudsters for a reasonable amount of time but the amount of attacks were largely limited by people with the knowledge and resources to create them.
Given the AI possesses the ability to generate complex code, it is possible that it could provide a user with a previously inaccessible weapon. A crude but accurate way to put it would be that it makes dumb people dangerous.
What Does ChatGPT Think?
Interestingly, when I asked ChatGPT to generate an article on its own risks, a primary concern was the use of it to “create harmful or inappropriate content such as hate speech or misinformation.” Given the ability of AI to produce content rapidly, the potential for harmful content to be used with malicious intent is highly concerning. It will be interesting to see how the necessary safeguards are implemented to protect users. Is it possible to filter everything or will parts inevitably slip through the cracks? Additionally what level of safeguarding is acceptable, is it 100%, 90%, 70%? These are important discussions that need to take place and standards must be set and enforced.
The Elon Effect
Finally, for those who have been paying attention to Elon Musk (a cofounder of OpenAI) and his barrage of tweets, you may have noticed he has distanced himself from the company.
OpenAI was created as an open source (which is why I named it “Open” AI), non-profit company to serve as a counterweight to Google, but now it has become a closed source, maximum-profit company effectively controlled by Microsoft.
Not what I intended at all.
— Elon Musk (@elonmusk) February 17, 2023
It seems that his opinion is the product is not reflective on the founding ideas of the company, stating it was “Not what I intended at all”. When it comes to the issue of privacy, what could be more incriminating than the criticism of a cofounder?
ChatGPT is undoubtedly an amazing innovative piece of technology, this cannot be understated. However, like any emerging piece of technology, particularly at the onset of its creation there are several risks that must be addressed and safeguarded otherwise the results could be catastrophic.
We would love to know what you think about ChatGPT so let us know what you think on our Twitter, Instagram, and Facebook.
For more information on emerging technology, checkout IVAS: What Is It and Why Is It Important? .
