Header image: KF in DALL-E
(Mar 1, 2024) Five key elements of Canada’s new Online Harms Act
https://srinstitute.utoronto.ca/news/canada-online-harms-five-key-aspects
- presented by Justice Minister Arif Virani on February 26, 2024, following an unsuccessful attempt at legislation in 2021.
- The bill proposes changes to the Criminal Code of Canada and the Canadian Human Rights Act, recommending stiffer sentences for hate speech related offenses and taking a new approach to combating hate speech by classifying it as discrimination under the Canadian Human Rights Act.
- However, most of the bill is devoted to the Online Harms Act, an important piece of new legislation that creates a new administrative apparatus, imposes new obligations on social media companies and online platforms, and addresses important issues like non consensual intimate imagery and safety for children online.
- The act creates three new institutions
- The Digital Safety Commission is tasked with administering and enforcing the act, plus investigating complaints.
- The Digital Safety Ombudsperson
- The Digital Safety Office provides administrative support to both the commission and the ombudsperson.
- The act proposes to regulate seven kinds of online harm
- Intimate content shared without consent
- Content that sexually victimizes a child or revictimizes a survivor
- Content that induces a child to harm themselves
- Content used to bully a child
- Content that foments hatred
- Content that incites violence
- Content that incites violent extremism or terrorism
- The act offers two methods for removal of harmful content
- Companies that don’t comply could face steep penalties, with fines of up to $10 million or six per cent of global revenue. Violating an order of the commission constitutes an offense attracting fines of up to $25 million or eight per cent of global revenue.
- The act creates new protections for children online
- The third harm is content used to bully a child, referring to content that can seriously impact a child’s mental or physical health, or otherwise threaten, intimidate or humiliate the child. It must be “reasonably suspected” that this kind of content was communicated for these purposes.
- The act addresses the unique harms posed by deepfakes and automated communication
- The bill also requires operators of social media services to label harmful content that has been repeatedly communicated by automated systems, like spambots.
However, some important questions about the bill remain unanswered.
- seeks to appropriately balance very real concerns about online harms with the need to uphold fundamental freedoms of speech and expression.
(March 14, 2024) EU Parliament passes AI Act in world’s first attempt at regulating the technology https://therecord.media/eu-parliament-passes-ai-act-regulation
- passed by a wide margin with 523 votes in favor, 46 against and 49 abstentions,
- aims to “protect fundamental rights, democracy, the rule of law and environmental sustainability from high-risk AI, while boosting innovation and establishing Europe as a leader in the field.”
- will go through a final check by lawyers and linguists before being formally endorsed by the European Council.
- categorizes AI systems by the potential harm they could do if they fail to work as intended or advertised, going from low-risk services such as spam filters through to high-risk tools used in critical infrastructure. The highest risk tier is “unacceptable” and these systems would be prohibited under the new law.
- France and Germany warning that the rules could hamper their domestic companies and potentially offer an advantage to competitors in the United States and China.
- The first impacts of the AI Act will be felt six months after it enters into force, when its provisions to protect fundamental rights — such as banning “biometric categorisation systems based on sensitive characteristics and untargeted scraping of facial images from the internet or CCTV footage to create facial recognition databases” — will apply.
- KF: presumably companmies will try to “make hay” in those six months
- Also prohibited will be the use of emotion recognition technologies in the workplace and schools, as well as social scoring and certain kinds of predictive policing based on profiling individuals.
- The use of real-time facial recognition systems by law enforcement is permitted “in exhaustively listed and narrowly defined situations,” when the geographic area and the length of deployment are constrained.
- KF: helpful restrictions that will inevitably be eroded over time
We finally have the world’s first binding law on artificial intelligence, to reduce risks, create opportunities, combat discrimination, and bring transparency. Thanks to Parliament, unacceptable AI practices will be banned in Europe and the rights of workers and citizens will be protected.
Brando Benifei, co-rapporteur
- A year and a day after the AI Act is published in the official journal of EU law, the rules around general-purpose AI and governance will enter into effect.
- These introduce obligations on developers to meet transparency requirements, including compliance with copyright law, and will force developers to be able to publish detailed summaries of the content they use for training.
- “We ensured that human beings and European values are at the very centre of AI’s development,” Benifei added.
AI will push us to rethink the social contract at the heart of our democracies, our education models, labour markets, and the way we conduct warfare”
Dragos Tudorache, co-rapporteur
See: Doctorow (2024) McLuhan lecture on enshittification https://kieranfor.de/2024/03/24/doctorow-2024-mcluhan-lecture-on-enshittification/