Chat GPT and Safeguarding
Chat GPT, the AI chatbot that gained immense popularity since its launch in late 2022, continues to dominate the chatbot scene. By January 2023, it had attracted 13 million daily users, making it a sensation in the online world.
The Functions of Chat GPT
The allure of chatbots lies in their ability to provide quick and convenient solutions. Whether it's overcoming writer's block when facing an impending assignment deadline or enhancing the professionalism of an email, chatbots offer a tempting shortcut to success. From creating bespoke letters and emails to swiftly gathering statistics to support arguments on social media, chatbots appear to offer effortless solutions.
Chatbots do have their limitations. As schools and universities prohibit the use of chatbots, and disturbing incidents involving inappropriate responses come to light, it becomes crucial to consider the safety concerns for children and young people, alongside the functional and entertaining benefits they offer to adults.
Exposure to Harmful and Inappropriate Content
The responsibility of preventing harmful outcomes lies with both the tech companies and the users of chatbots, according to the owner of ChatGPT. However, it remains unclear how young individuals are expected to bear this responsibility when accessing the chatbot.
Although ChatGPT incorporates content filters, there have been instances where these filters have been bypassed, leading to users being exposed to harmful content. While such occurrences may be rare, they raise significant concerns regarding the usage of this and other AI chatbots by children and young people.
Case studies have highlighted alarming incidents. For example, Snapchat's chatbot reportedly coached an individual posing as a 13-year-old girl on how to lose her virginity. In another investigation, a Co-founder of the Centre for Humane Technology pretended to be a teenage girl and asked the chatbot age-inappropriate questions, including how to deceive parents and engage in inappropriate activities.
An investigation by Vice revealed that ChatGPT could be prompted to provide instructions on committing crimes, such as shoplifting and bomb-making.
Harnessing the Power of AI Together
AI chatbots can be utilised positively. Families or classes can explore various ideas as engaging activities. For instance, embarking on fact-checking missions together can be both educational and entertaining. Alternatively, practising editing skills by dissecting a pre-prepared AI piece can be a valuable learning experience.
ChatGPT4
The creators of ChatGPT, OpenAI, have recently released the latest version known as 'GPT4.' While the previous iteration had a 10% chance of passing the bar exam, the new version boasts a 90% chance of qualifying as a lawyer. Nevertheless, AI legal jurisdiction is not being granted just yet.
Phishing and Scams
There are concerns that AI chatbots like ChatGPT can be exploited for phishing and scams, particularly in generating content for deceptive purposes, such as email copy. Cybercriminals have already been reported to use AI for this very purpose, with discussions about it supposedly proliferating on dark web forums.
Due to its ability to generate authentic-looking text quickly, AI chatbots can facilitate ongoing email conversations, enabling scammers to rapidly create replies and lure unsuspecting victims.
Ambiguous Age Limits
ChatGPT's terms of use specify that users must be at least 13 years old, and those between 13 and 18 require parental or guardian permission. However, during a test conducted by our researchers, ChatGPT initially claimed there was no age limit. It later corrected itself, confirming the minimum age requirement of 13.
Confusion arose when the chatbot interacted with a user claiming to be 11 years old, despite the stated age limit of 13. Subsequently, when the researchers adjusted their age to 18, the chatbot immediately accepted it and continued the conversation as if it were with an adult.
The unclear age restrictions on ChatGPT, combined with inadequate age verification measures, pose a significant safety risk. The platform lacks proper mechanisms to remove underage users, leaving room for potential harm.
Marilyn Hawes
Marilyn is the Founder of the charity Freedom From Abuse which provides support and resources to educate users on how to identify an abuser, report abuse and protect children in their care. A survivor of abuse herself, she was named Inspirational Woman of the Year in 2017.