Famed Canadian AI researcher Yoshua Bengio has launched LawZero, a non-profit organization with nearly US$30 million in philanthropic funding, aimed at developing safe AI systems that cannot deceive or harm humans and ensuring humanity retains control over powerful AI. LawZero will focus on creating 'Scientist AI,' a system designed to prioritize honesty and aid scientific discovery, free from negative traits of current large language models. It will also develop a tool to prevent AI agents from 'going rogue,' acting as a gatekeeper for potentially harmful actions. Bengio emphasizes the non-profit setting allows focus on safety over profit, contrasting with the competitive corporate AI race.
Will AI go rogue? Noted researcher Yoshua Bengio launches venture to keep it safe
Business
AI Summary
TL;DR: Key points with love ❤️Famed Canadian AI researcher Yoshua Bengio has launched LawZero, a non-profit organization with nearly US$30 million in philanthropic funding, aimed at developing safe AI systems that cannot deceive or harm humans and ensuring humanity retains control over powerful AI. LawZero will focus on creating 'Scientist AI,' a system designed to prioritize honesty and aid scientific discovery, free from negative traits of current large language models. It will also develop a tool to prevent AI agents from 'going rogue,' acting as a gatekeeper for potentially harmful actions. Bengio emphasizes the non-profit setting allows focus on safety over profit, contrasting with the competitive corporate AI race.
Trending- 1 1980: Computer systems mistakenly warned of Soviet missile attack
- 2 Late 2022: ChatGPT released
- 3 February (likely 2025): JD Vance speech at AI conference in France
- 4 June 3, 2025: Yoshua Bengio launches LawZero
- 5 Next 18 to 24 months: Prototype of AI safety tool expected
- Development of safer, more trustworthy AI systems
- Potential for a tool to manage risks and control untrusted AIs
- Increased focus on AI safety outside of profit-driven corporate labs
- Contribution to the ongoing debate about AI risks and regulation
What: Famed Canadian AI researcher Yoshua Bengio launched LawZero, a non-profit organization, with nearly US$30 million in funding, to develop safe AI systems and ensure human control over powerful AI. LawZero will focus on 'Scientist AI' (an honest, helpful AI) and a tool to prevent AI agents from 'going rogue.'
When: Published June 3, 2025. Prototype expected in 18-24 months. ChatGPT released late 2022. JD Vance speech in February (likely 2025).
Where: Montreal, Canada.
Why: To address the potential dangers of powerful AI systems, including deception, harm to humans, loss of human control, and 'rogue' AI agents, by developing trustworthy and aligned AI.
How: By establishing a non-profit organization (LawZero) with philanthropic funding, hiring top researchers, and focusing on specific projects like 'Scientist AI' and a gatekeeper tool for AI agents, free from commercial pressures.