LawZero
A non-profit AI safety research organization co-founded by Yoshua Bengio, focused on researching and building AI systems capable of operating under constitutional constraints. Its goal is to achieve AI safety and controllability at the foundational architectural level.
Core Facts
Timeline (last 90 days)
This disclosure comes from a single social media source, and specific details regarding technical approaches, funding scale, and research outcomes have not been fully disclosed.
The presentation topic was titled "Engineering honesty and reliability for safer AI."
LawZero argues that as frontier models become more powerful, two parallel lines of defense are needed: governance mechanisms at the societal level and robust guardrails at the technical level.
Relevant principals of LawZero shared its research agenda and recently secured funding in a keynote address at the ALL IN conference.
AI research institution LawZero announced that it has received support from the governments of both Canada and Germany.
The tweet has not yet disclosed the specific funding amount, details of research directions, or timeline for LawZero.
LawZero was founded in 2024, its full name is Law Zero for AI Safety, and it is a nonprofit AI safety research institution headquartered in Montreal.
Bengio will appear on the Canadian television program Tout le monde en parle to discuss LawZero's government support and AI safety incidents.
LawZero focuses on research into verifiable AI alignment methods, monitoring mechanisms, and policy frameworks.
LawZero, the nonprofit institution founded by Bengio, has received support from the governments of both Canada and Germany.
18 more timeline events
All Facts (17)
LawZero is a research organization focused on AI safety and governance, whose philosophy pursues a 'zero harm' baseline in the deployment of AI systems, with its focus on building trustworthy, controllable, and verifiable AI systems
80%UnverifiedLawZero's research involves transnational collaboration between Canada and Germany.
70%UnverifiedLawZero argues that as frontier models become more powerful, two parallel lines of defense are needed: governance mechanisms at the societal level and robust guardrails at the technical level.
50%UnverifiedLawZero focuses on research into verifiable AI alignment methods, monitoring mechanisms, and policy frameworks.
50%UnverifiedLawZero was founded in 2024, its full name is Law Zero for AI Safety, and it is a nonprofit AI safety research institution headquartered in Montreal.
50%UnverifiedThe research path toward fundamental honesty attempts to endow models with the metacognitive ability to know what they do not know—through training objectives and architectural design—thereby reducing hallucinations.
50%UnverifiedLawZero is dedicated to designing AI systems that are fundamentally honest and trustworthy, treating credibility and honesty as the foundational starting point of system design.
50%UnverifiedLawZero is a nonprofit organization founded in 2024 and co-founded by Yoshua Bengio.
50%UnverifiedLawZero's core objective is to research and build AI systems capable of operating under constitutional constraints.
50%UnverifiedThe organization LawZero is working on embedding safety properties at the foundational level of model training.
50%UnverifiedAccording to Bengio's statements, LawZero is in a phase of rapid growth, actively expanding in resource investment, talent recruitment, and research directions.
50%UnverifiedThe name LawZero echoes the Zeroth Law in Asimov's laws of robotics, which states that a robot may not harm humanity as a whole.
50%UnverifiedLawZero is a nonprofit research organization that Yoshua Bengio recently spearheaded the establishment of.
50%UnverifiedThe Scientist AI approach includes explicit honesty constraints, belief coherence evaluation, and a transparency-first principle.
50%UnverifiedLawZero is exploring the Scientist AI approach, placing honesty and internal belief coherence at the core of the training process.
50%UnverifiedLawZero, the nonprofit institution founded by Bengio, has received support from the governments of both Canada and Germany.
50%UnverifiedResearch institution LawZero is rethinking AI system training methods from a more fundamental level.
50%