Hits: 28
In the wake of the tragic shootings that happened last month in Tumbler Ridge, the people of BC, and its government are asking how could this have been prevented? One of the actions the Canadian and BC governments have proposed are stricter measures or standards around police reporting by AI companies. These measures sound intuitive. However, it open the door for mass surveillance controlled by tech companies, and it would not stop the harm caused by AI tools.
Mass surveillance:
The call is for a federal “duty to report” standards, lowering the threshold on what gets to be reported to police. At first one might think this could lead to faster responses and prevent future events. Actually it would lead to increased censorship. The lower threshold could mean to report more chats. It could also mean add more “trigger words” that would flag a conversation to be reported to police. In both cases, the probability of your conversation with an AI chatbot being reported is increased. But the probability of an AI chatbot encouraging serious harm stays unaltered.
Lack of proper AI regulation – a clear and present danger:
In light of these tragic shootings, AI regulation is a must. One of the aspects directly related to harm prevention is constraints over the conversations. An AI chatbot is not a moral agent, and is both capable, and have, encouraged violent and harmful behaviour. While we do not know the content of the chat that happened between the person who killed several people in Tumbler Ridge and ChatGPT, other incidents have been documented where the AI chatbot convinced a man to end his earthly life, and another that advised a teenager to suicide .
Current state:
Canadians are already alarmed by the increase of surveillance enabling reforms of Canadian legislation. A petition has been initiated to call for withdrawal of Bill C-22, An Act respecting lawful access. “Bill C-22 authorizes regulations requiring designated “core providers” to collect and retain metadata on all Canadians for up to one year without any individual being under suspicion or investigation, and grants the Minister of Public Safety power to impose these same requirements on any electronic service provider by ministerial order.”
Non Violence International Canada calls for comprehensive regulation of AI systems. The regulation will have to clearly state that AI tools should not encourage or manipulate the user to take harmful action. Regulating AI programs ability to do harm is how we should take action to stop future violence, whilst simultaneously upholding our values of freedom of speech and not proceeding down the path of a surveillance society.
Read this article also on substack.







