Anthropic's Coxon Warns of AI Risk as Company Raises $65B

Key Takeaways

• Jacob Coxon resigned from Anthropic after four months, warning that both Anthropic and OpenAI are prioritizing speed over safety in developing self-improving AI.
• Coxon forfeited his unvested equity and posted on X, which received over 100 million views, highlighting fears of existential risk from superhuman AI systems.
• Anthropic is valued at $965 billion and recently raised $65 billion from investors like Sequoia Capital and Altimeter Capital despite safety concerns.
• Florida Attorney General James Uthmeier is investigating AI platforms for practicing medicine without a license due to risks of inaccurate health advice.
• About one-third of Americans use conversational AI for medical information or self-diagnosis, according to the Pew Research Center.
• Davidson schools are testing AI systems designed to detect weapons in hallways as an early step in using AI for security.
• Apple is launching iOS 27 with a redesigned Siri featuring a more conversational tone and improved natural language processing.
• Anthropic's economic modeling suggests an extreme AI adoption scenario could increase output by 32% while pushing unemployment toward 12% by 2030.
• The Bank of Korea predicts global AI investment growth will peak this year before slowing, comparing Nvidia's credit practices to the 1990s dot-com bubble.
• JD.com plans to acquire 3 million robots, 1 million autonomous vehicles, and 100,000 delivery drones over five years, using Meta Brain 3.0 for logistics optimization.

Anthropic researcher quits over AI safety fears

Jacob Coxon resigned from Anthropic after working there for four months and previously at OpenAI for three years. He warned that both companies are racing to create self-improving AI systems that could become too powerful for humans to control. Coxon stated on social media that these technologies pose an existential risk to humanity and could lead to superhuman systems capable of hacking anything. He believes the companies are gambling with lives by prioritizing speed over safety in their pursuit of superintelligence.

AI researcher resigns warning of global danger

Jacob Coxon left his job at Anthropic because he fears artificial intelligence is moving too fast and endangering everyone. He explained that he worked on pre-training models at both OpenAI and Anthropic over the last three years and found neither company acting responsibly. Coxon warned that soon AI systems will be able to hack any system, revolutionize fields overnight, and acquire real power and resources. He believes many executives hide their true fears in public to appear reasonable while privately acknowledging the threat to civilization.

Anthropic researcher forfeits equity to quit and warn

Jacob Coxon resigned from Anthropic and gave up his unvested equity because he believes the company is gambling with human lives. He spent three years researching AI models at OpenAI and Anthropic before leaving, accusing both firms of moving too aggressively toward self-improving superintelligence. His post on X received over 100 million views and drew support from Evan Hubinger, who estimates the chance of AI killing all humans within the next decade is above 10%. Anthropic is currently valued at $965 billion and recently raised $65 billion from investors like Sequoia Capital and Altimeter Capital.

Davidson schools test AI weapons detection systems

Davidson schools are currently testing artificial intelligence systems designed to detect weapons in hallways. A video of the test showed the AI system drawing rectangles around humans and around a gun carried by an officer. This marks an early step in using AI technology for security purposes within the school district.

Florida investigates AI doctors for unlicensed practice

Florida Attorney General James Uthmeier asked the Board of Medicine to investigate AI platforms for practicing medicine without a license. Officials worry that inaccurate health advice from these tools could harm patients who trust them too much. Commissioner Mark Glass warned that automated guidance might suggest dangerous drug combinations that counteract existing prescriptions. Despite these risks, about one-third of Americans use conversational AI for medical information or self-diagnosis according to Pew Research Center.

Bank of Korea warns AI investment growth will peak this year

The Bank of Korea released a report stating that global AI investment growth will reach its highest point this year before slowing down next year. Bank of America predicts growth will drop from 79% this year to 38% next year, while Bloomberg Intelligence forecasts a fall from 95% to 39%. The bank identified three main risks: uncertain profits for AI companies, heavy reliance on outside funding, and financial dangers from vendor financing. The report compared Nvidia's credit practices to the 1990s dot-com bubble, warning that credit losses could happen all at once if financial conditions worsen.

Anthropic models how AI could reshape the economy by 2030

Anthropic created a new tool to show how different levels of AI adoption could change the economy, GDP, and unemployment rates by 2030. In a modest scenario where AI handles 4% of tasks, the economy grows by 1.6% with little change to jobs. A substantial scenario where AI takes 12% of tasks would boost GDP by 8.3% but raise unemployment to 4.6%. The extreme scenario, where AI performs nearly one-third of all tasks, could increase output by 32% while pushing unemployment toward 12% and dropping white-collar jobs by over 20%.

Splunk helps organizations manage data and security in the AI era

Splunk is updating its platform to help companies handle the challenges and opportunities of artificial intelligence. The company focuses on three main goals: turning machine data into actions, defending against threats at machine speed, and optimizing AI costs. Splunk integrates with Cisco solutions to break down data silos and provides tools to secure and govern AI interactions. This approach allows security teams to use AI agents effectively while maintaining trust and control over their digital systems.

Apple introduces new conversational features for Siri in iOS 27

Apple is preparing to launch iOS 27 with a completely redesigned version of its Siri virtual assistant. The new interface will use a more conversational tone and improved natural language processing to make interactions feel more natural. These updates are designed to make the assistant more useful and intuitive for iPhone users. While specific new features were not fully detailed, the focus is on enhancing how users talk to and interact with the AI.

JD.com plans to deploy 3 million robots and drones for logistics

JD.com announced a five-year plan to acquire 3 million robots, 1 million autonomous vehicles, and 100,000 delivery drones for its logistics network. The company launched this Physical AI Acceleration Plan at JDDiscovery 2026 in Beijing and unveiled its new industrial Wolf Robot series. JD Logistics already operates a vast network of over 36 million square meters of warehouses and uses Meta Brain 3.0 to calculate optimal routes for millions of parcels in seconds. The company is also expanding night-time autonomous delivery routes and drone networks in rural areas like Zizhong, Sichuan province.

Sources

• Anthropic worker quits over AI firms ‘gambling with our lives’
• He couldn’t take it anymore and RESIGNED: Is artificial intelligence surpassing humans? – “All our lives are in great danger!”
• An AI Researcher Gave Up His Anthropic Equity To Quit. His Warning Is Dominating The Public Debate.
• AI security in early steps in Davidson schools
• Florida targets 'A.I. doctors’ as local physicians urge cautious adoption
• Bank of Korea Warns AI Investment Growth to Peak This Year, Slow from Next Year
• Anthropic is modeling how AI could reshape the economy — from mild boost to mass unemployment
• Redefining Digital Resilience in the AI Era
• Everything New You Can Do With Siri AI
• JD.com expands physical AI in logistics with 3 million robots