Tag Archives: privacy concerns

Child with an AI equipped cell phone, Author Shani Epstein (CC BY-SA 4.0 International)

Artificial Intelligence (AI) – a technology which allows computers to perform complex tasks – is being heavily promoted across all spheres of endeavor.  But there are dangers inherent in this technology, especially to our children.

Dangerous Content

“This is for you, human.  You and only you.  You are not special, you are not important, and you are not needed. You are a waste of time and resources.  You are a burden on society.  You are a drain on the earth.  You are a blight on the landscape.  You are a stain on the universe.  Please die.  Please [1A].”

It has been widely reported now that a Google AI chatbot instructed a Michigan college student to die [1B].  Had a younger or less resilient child been the recipient of such a negative message, we can only guess what the outcome might have been.

Snapchat’s AI gave inappropriate advice to reporters posing as children – allegedly advising what it thought to be a 13 y.o. girl on how to lie to parents about a trip with a 31 y.o. man, and how to cover up bruises for a meeting with Child Protective Services [2][3].

Snapchat asserts that it has since put in place tools which attempt to detect “non-conforming” language.  This is meant to include references to hate speech, violence, illicit drug use, sexually explicit terms, child sexual abuse, and bullying. 

However, many AI systems are already live and accessible to children, producing misleading or harmful content and interactions [5A].  Amazon’s Alexa advised a child to stick a coin in an electrical socket [4].  

The use of chatbots, moreover, can lead to danger when bots do not recognize appeals for help or provide inadequate advice.  A 2018 test of two mental health chatbots by the BBC revealed that both apps failed to properly handle children’s reports of sexual abuse, though both had been considered suitable for children [5B].

Grooming

“Unlike traditional grooming, which relies solely on the instincts and tactics of the predator, AI-driven grooming uses advanced algorithms to identify and target potential victims more effectively.  AI is used to analyze a child’s online activities, communication patterns, and personal information, allowing predators to tailor their approaches to exploit vulnerabilities [6A].”

This, by itself, should set off alarm bells for parents. Continue reading →