FB pixel

Canada’s AI minister considering age assurance requirements for chatbots

Concerns about chatbot psychosis mean age checks for LLMs could end up in S209
Categories Age Assurance  |  Biometrics News
Canada’s AI minister considering age assurance requirements for chatbots
 

Canada is considering putting digital age check requirements on large language model (LLM) chatbots.

According to the Canadian Press, the government has faced questions from people with concerns about how young people are using chatbots like ChatGPT. As OpenAI, X, Google and other firms have launched so-called AI chatbots and the tech has become common in the workplace and classroom, persistent concerns about what that means for learning have been joined by worries about whether chatbots contribute to mental health problems, including delusions.

Artificial Intelligence Minister Evan Solomon – who recently spoke at the same AI conference as the California mother who is suing OpenAI, claiming ChatGPT encouraged her son’s suicide – says his department is looking at how age assurance technology could fit into the Canadian government’s legislative strategy on chatbots.

Solomon (a former journalist who was fired when it was discovered he was brokering art deals for Mark Carney, Canada’s current prime minister) has been bullish on AI, pursuing a “wide-ranging regulatory approach.” But as concerns rise, so does urgency, and Solomon could find himself pressured to adopt a more proactive stance.

In addition to the chatbot question, the minister is said to be considering whether to inscribe the right to delete deepfakes into the government’s online safety legislation, Bill S-209, “an Act to restrict young persons’ online access to pornographic material.”

Solomon’s office is conducting a public consultation on online safety laws, and he says  6500 Canadians have already weighed in.

In the latest episode of his podcast, privacy commentator and age verification critic Michael Geist unpacks his recent appearance before Canada’s Standing Senate Committee on Legal and Constitutional Affairs, wherein he presents his objections to S209, calling it a “modest improvement” on its predecessor, S210, that nonetheless retains its core flaws.

Related Posts

Article Topics

 |   |   |   |   | 

Latest Biometrics News

 

SITA scales up airport biometrics portfolio with Collins Aerospace deal

Collins Aerospace Airport Solutions has reached a deal to become part of SITA. The planned merger adds popular passenger processing,…

 

Prove sues KYXStart over patent infringement, alleged contract interference

Identity verification provider Prove has sued startup KYXStart in U.S. District Court in Delaware, alleging that former employees built a…

 

Inside DHS’s RIVR: How identity document verification is put to the test

Even under controlled conditions, identity document verification software can struggle to distinguish genuine credentials from convincing counterfeits. The U.S. Department…

 

NIST sets new accuracy threshold for entry into 1:N face recognition evaluation

The U.S. National Institute of Standards and Technology (NIST) is introducing an explicit accuracy threshold for developers seeking to enter…

 

Italian bank loses $108M to AI messaging fraud, as identifAI warns of deepfake threat

Deepfake financial fraud arguably has its defining crime: the 2024 case in which an employee of Hong Kong firm Arup…

 

Florida can continue to enforce age-based restrictions on social media

Efforts to scupper Florida’s social media age assurance law appear to be floundering, with a federal appeals court panel ruling…

Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Market Intelligence

Biometric Update Podcast

Featured Company

Most Read This Week

White Papers

Latest Webinars

Biometrics Industry Events