August 2026

This monthly newsletter covers recent developments and upcoming events in AI safety, ethics, and governance in Montréal.

Events

AI Safety Papers We Love #4 - Law-Following AI: Designing AI Agents to Obey Human Laws
Tuesday, August 11, 6 to 8 PM. Ω Labs, 3813 rue Saint-Denis, Montréal.
Talita Dias, an international lawyer and Professor of Practice at McGill (previously Partnership on AI, Chatham House, and Oxford), presents “Law-Following AI” by O’Keefe, Ramakrishnan, Tay, and Winter: in high-stakes settings like government, AI agents should be designed to rigorously follow the law, loyal to their principals but only within legal bounds, with legal duties imposed directly on AI agents.

Introducing ControlAI Canada
Thursday, August 13, 6 to 8 PM. Ω Labs, 3813 rue Saint-Denis, Montréal.
ControlAI is a global advocacy nonprofit working to prevent the extinction risk posed by superintelligence. Étienne presents its work worldwide and in Canada, followed by a write-to-your-MP activity.

AI Safety Speed-networking & social
Tuesday, August 18, 6 to 8 PM. Ω Labs, 3813 rue Saint-Denis, Montréal.
A relaxed evening for anyone working on, or curious about, AI safety, ethics, and governance in Montréal: a few rounds of speed-networking, then open socializing.

IVADO thematic semester: The Governance of AI
August to October, IVADO and HEC Montréal. Bilingual, registration open.
A semester-long program with CNRS on multi-level AI governance. Workshop 1, “Governing AI: Scale, Scope and Approach,” runs Monday, August 24 to Thursday, August 27. Workshop 2 on sector-specific governance follows September 14 to 18, and a September 18 conference on interdisciplinary AI governance.

Canada AI Transparency Survey Jam
Tuesday, August 25, 6:30 to 8:30 PM. Ω Labs, 3813 rue Saint-Denis, Montréal.
A hands-on evening to answer ISED’s AI transparency consultation together (see Opportunities below). Bring a laptop and leave with your response submitted.

AI Governance Reading Group #1 - What AI evaluations for preventing catastrophic risks can and cannot do
Thursday, August 27, 6 to 7:30 PM. Ω Labs, 3813 rue Saint-Denis, Montréal.
First session of a new reading group on AI governance. We start with Barnett and Thiergart: what evaluations of frontier models can actually establish about catastrophic risk, and what are their limits. Advance reading encouraged.

Opportunities

Federal consultation on AI transparency, open until September 23
ISED is consulting on consistent public information about AI systems, tracking of serious AI incidents, and ways to track AI agents’ activities and interactions.

OBVIA: call for AI workshop facilitators, deadline September 4
OBVIA is recruiting paid facilitators to run awareness and reflection workshops on AI issues for varied audiences.

Mila AI Policy Compass, French cohort September 1 and 2
Twelve hours over two days at Mila’s Montréal offices, run with the Public Policy Forum, for policy and decision makers working through AI opportunities and risks.

IVADO R3AI research clusters 2027–2030: full proposals due November 19
Up to $20M for 8 to 12 interdisciplinary clusters under the Robust, Reasoning and Responsible AI program, across UdeM, Polytechnique, HEC, Laval, and McGill.

Mila supervision requests open October 15
The annual window for reaching out to potential supervisors at Mila.

Policy and Governance

OBVIA files a brief on AI in Québec public services
July 15.
OBVIA submitted recommendations to the Protecteur du citoyen’s investigation into the use of AI and algorithms in Québec public services, announced in May and collecting briefs through July 15.

B.C. prepares a lawsuit against OpenAI over the Tumbler Ridge shooting
July 7.
British Columbia’s attorney general says the province has retained counsel to sue OpenAI over its alleged role in the February mass shooting. The case, and parallel federal consultations, will shape duty-to-report norms for AI companies operating in Canada.

CAISI asks what information AI evaluators should share
July 8.
The Canadian AI Safety Institute’s first substantive post, with the AAIMES Network, proposes reporting norms for model evaluations: disclose intent, methodology, model specifications, limitations, and conflicts of interest, while withholding adversarial techniques and safety critical datasets.

Bill C-36: Canada’s third attempt at privacy reform, with automated-decision transparency
At second reading in the House of Commons. Parliament resumes September 21.
The Protecting Privacy and Consumer Data Act would replace PIPEDA, treat children’s data as sensitive, require transparency for automated decisions, and create a Digital Safety and Data Protection Commission with penalties up to $25M or 5% of global revenue.

Bill C-34: the Safe Social Media Act reaches AI chatbots
At second reading in the House of Commons. Parliament resumes September 21.
Beyond social media age limits for under-16s, the bill would impose safety obligations on AI-powered chatbot services, with fines up to 3% of global revenue, and create a Digital Safety Commission to enforce them.

Research from Montréal

Bengio co-chairs the UN scientific panel’s first report on AI
July 7.
Yoshua Bengio (UdeM, Mila) co-chairs the UN’s independent scientific panel on AI with Maria Ressa, and the panel published its preliminary report on July 1. Its 40 members were appointed by the General Assembly in February and elected their co-chairs in March.

LawZero gets compute from NVIDIA for its Scientist AI
July 27. LawZero, Montréal.
NVIDIA has donated high-performance computing to LawZero, the Montréal nonprofit led by Yoshua Bengio, for training and evaluating the core components of its Scientist AI at scale. Co-president and executive director Sam Ramadori calls the contribution instrumental to the science behind the design. It follows the formal safety case the team published in July.

ROOST and Mila pilot a suicide-prevention guardrail for chatbots
July 8.
An open-source pilot detecting chatbot responses that could encourage self-harm in young users, the first release from Mila’s AI Safety Studio, landing exactly where Bill C-34’s chatbot provisions are aimed.

Arts and Culture

MUTEK Forum 2026
Wednesday, August 26 to Friday, August 28. Grands Ballets and Édifice Wilder, Montréal.
This year’s Forum addresses the relationships between AI, living systems, artistic practices, and technological infrastructure, toward more ethical, sustainable, human-centred technological futures. Day 2 focuses on AI, generative systems, and human-machine collaboration. (The MUTEK festival itself runs August 25 to 30.)

“À l’heure de l’IA”, final month at the Grande Bibliothèque
Through Sunday, August 30. Grande Bibliothèque (BAnQ), 475 boul. de Maisonneuve E, Montréal. Free.
OBVIA’s interactive exhibition on the societal impacts of AI, eight everyday-life scenes plus modules on the ethical principles of responsible AI, closes at the end of August.

Have an event, opportunity, or article for next month? Send us a note at team@horizonomega.org.

This newsletter is by HΩ, researched and written with the assistance of AI.