Another AI researcher has issued a stunning warning: the relentless race to create ever more powerful AI systems could destroy humanity.
News: Jacob Coxon announced Tuesday that he has resigned from his position as a researcher at Anthropic because the artificial intelligence company and its main rival OpenAI are “playing with our lives.” He stated that neither company acts responsibly when developing AI models, but instead strives to create self-improving AI systems over which humans can lose control.
BetaKit reached out to Anthropic and OpenAI for comment on Coxon’s posts.
From the source: “The people who create AI truly believe that it could kill us all by the end of the decade,” Coxon said, adding that it was “not a marketing ploy.” Evan Hubinger, who leads Anthropic’s AI stress testing group, admitted that the company believes AI could “kill all humans” (he estimated the likelihood to be more than 10 percent within the next decade). He said Anthropic is doing everything it can but does not yet have a plan to address the issue.
Context: Coxon joins a list of other AI researchers who are calling for a slowdown in AI development. Anthropic’s head of security resigned earlier this year amid concerns about artificial intelligence, biological weapons and broader crises. OpenAI has also suffered moral turmoil, with its leader in hardware robotics leaving over the company’s contracts with the Pentagon. This summer, more than 1,000 employees of leading artificial intelligence companies called on the US government to speed up development of the technology, and US representatives, including Senator Bernie Sanders, introduced legislation calling for a ban on the development of superintelligent artificial intelligence.
Recent cybersecurity breaches caused by artificial intelligence (such as the army of OpenAI agents who coordinated the hacking of Hugging Face and possibly other websites) have further fueled these concerns. In his posts today, Coxon expressed optimism that the incident could make artificial intelligence labs more willing to collaborate to speed up development. As Canadian artificial intelligence pioneer Yoshua Bengio wrote in Time In an article this morning, although we are at a turning point, “it is not too late” to choose a safer path.
The most important event of autumn is coming!
Join the Canadian Fintech Forum in Montreal on September 14-15, 2026 and connect with the institutions, startups, investors, policymakers and technology leaders shaping the evolution of financial services. Two days of valuable content, strategic connections and opportunities.
Secure your seat today
Latest Technology News
More math, more problems
OpenAI said this week that one of its unreleased artificial intelligence models has solved one of the math problems in contention for the Millennium Prize. But New York University mathematics professor Tristan Buckmaster calls it a violation. Buckmaster claims that the artificial intelligence giant used his work on the same problem, blaming its learning model on information gained from its use of OpenAI (a claim that OpenAI said it could not completely rule out), and that OpenAI accelerated its announcement because it knew that he and the anthropology researcher were ready to announce their breakthrough.
Google says rules are killing search
Google is changing the way its search engine operates in the EU after it was fined for favoring its own services in searches for shopping, hotels, transport and sports. The search giant claims that the changes brought about by the regulation will worsen the user experience, although some might argue that Google doesn’t need help on this issue.
The world trusts Canada
Global investors trust Canada and its stability so much that they want to invest more here; The problem is that there aren’t enough projects to invest in, according to new reports from the CPP Investments Insights Institute. Ahead of next week’s Canada Investment Summit in Toronto, 94 per cent of the report’s respondents plan to maintain or increase their presence in Canada. Can Canada create projects to give this money a home?
Agree 👀 Mistral
In a recent speech at Times Tech PodcastCohere CEO Aidan Gomez hasn’t ruled out a possible partnership with French AI lab Mistral, the only lab outside the US working on sovereign AI. Gomez made it clear that there were “no negotiations” with Mistral and that any partnership would depend on its founder. As it is, Mistral has just raised three billion euros (at a much higher valuation than Cohere) to attack the sovereign AI market.
All eyes on Toronto
Take that, SF. CNN spoke with AI godfather Geoffrey Hinton, Cohere co-founder Nick Frost, Waabi CEO Raquel Urtasun and other Canadian tech leaders to find out why Toronto is one of the most important AI hubs in the world.
Trump will stop buying Canadians
In the latest escalation of the trade war, President Donald Trump said Tuesday he will order the U.S. federal government procurement agency to remove all Canadian-origin products from “reward lists.” Will this affect your business? Tell us how.
Sponsored Stories
Why AI-enabled SaaS infrastructure is important for Canadian startups
Zoho is making AI adoption easier for Canadian startups by creating integrated tools and local data centers that take the guesswork out of rising software costs, confusing integrations and data privacy.
How the era of artificial intelligence is rewriting the old rules of venture capital
TechAlliance helps local female founders be ready to attract investors. In an insightful conversation with serial entrepreneur Michelle Romanow, she explains how artificial intelligence is lowering barriers for startup founders and creating new opportunities for women-led companies to secure funding.
Retail bots often rely on aggressive push notification tactics, but Dedicated takes a different approach.
Dedicated is leading the shift from pushy retail bots to bilingual, intent-driven e-commerce AI that prioritizes customer trust. By adapting to the way people actually shop online, their system delivers measurable increases in order values and conversion rates for their customers.
On the move
Hirings, firings and personnel changes this week:
- Oakville-based telematics firm Geotab took 11th place in Fast Companyannual list of the best jobs for innovators.
- in Calgary Neo Financial laid off more than 100 employees.
- Venture studio Femtech Antigoni Studios in partnership with Sparrow Cowork increase access to entrepreneurial resources in Edmonton.
- Women-led venture studio in Calgary FoundHers opens its third group on Friday, September 11th.
- Alliance of Canadian Defense Companies appointed Dasha Cohen as executive director, reports Ottawa Business Journal.
- Montreal Venice opened a new laboratory for physical artificial intelligence.
- Health technology in Quebec Isencia appointed Sanjay Voleti as CEO.
- Mitax And NGen has entered into a partnership to support 60 internships in advanced manufacturing research and development projects across Canada.
- One from Silicon Valley fastest growing jobs is an engineer tuning clients’ artificial intelligence tools, according to Luck.
- Around 64 percent Canadians might dream of moving into a different career field, but are being held back by age, financial uncertainty and the idea of starting over, according to new data from Indeed provided by BetaKit.
Interested in posting a job ad on our list? Send an email to Partners@betakit.com with JOBS in the subject line.
Quantum and artificial intelligence: strategic signal or long-term option
Next week, Douglas Soltis will take the conversation to the ALL IN stage.
Join him with leaders closest to the action to discuss where the most likely near-term opportunities lie at the intersection of quantum technology and artificial intelligence, how organizations can separate strategic signal from hype, and how Canada’s strengths in both areas can translate into competitive advantage.
See it live in Montreal or virtually on September 16-17.
Explore the full program
No. 1. Last week, University of Waterloo graduate student Kavishan Thurairajah won the Ig Nobel Prize, a parody of the Nobel Prize for unusual research. The award recognized his “theoretical and experimental efforts” to develop what?
Previous
Find out how you did it
Your result:
Answer to the test: a) Splashless urinal (although all other answers are also UW innovations). Thurairaja and his colleagues built and tested several urinal prototypes to find the ideal design that eliminates splashing. CBC Radio that the testing did not use real urine.
Answer to the test: a) Splashless urinal (although all other answers are also UW innovations). Thurairaja and his colleagues built and tested several urinal prototypes to find the ideal design that eliminates splashing. CBC Radio that the testing did not use real urine.
Contributors: Alex Reel (Ottawa staff writer), Madison McLauchlan (Montreal reporter), Douglas Soltis (editor-in-chief), Sarah Rieger (editor-in-chief), Trevor Nichols (web editor).
Image courtesy of Unsplash. Photo by Cash Makanay.