Emi Yoshikawa, Ripple’s former vice president of global operations, has weighed in on a public resignation that is moving through tech and investment circles. On X, she shared a post about Jacob Coxon, an AI researcher who left Anthropic after spending three years on pretraining work at OpenAI and Anthropic. She quoted an insider who said the resignation letter was spreading fast. The same insider added that developers inside the company genuinely believe artificial intelligence could destroy humanity within a few years.
A resignation that traveled fast
Coxon did not leave quietly. In his statement, he accused leaders at both labs of running an irresponsible race toward self-improving superintelligence. He described the work as a gamble with human lives. His claim is not just about one company. It points to a wider concern inside the AI field: the people building advanced models may understand the risks better than the public does.
According to Coxon, the two labs handle that risk differently. Many OpenAI employees still do not fully grasp the scale of the danger, he wrote. At Anthropic, he said, researchers recognize the threat but keep moving because they fear losing the competitive race. That distinction matters. It suggests that awareness alone has not slowed development.
Why fintech and VC circles are paying attention
Yoshikawa’s response adds another layer. She is no longer only a Ripple executive. She now moves through venture capital projects and international investment networks. When someone with that background amplifies a warning from inside an AI lab, the message reaches people who think about risk in financial terms.
For fintech investors, internal turmoil at AI companies is no longer just a cultural story. It can become a valuation issue. If senior researchers publicly question the safety of their own work, that can shake confidence ahead of new funding rounds. The effect may be stronger than a regulatory forecast, because it comes from people who helped build the systems.
The valuation risk behind internal warnings
Anthropic and OpenAI both face pressure to ship better models quickly. That pressure does not disappear when an employee resigns. But a resignation letter from a pretraining specialist can still damage reputation and complicate fundraising. Investors may ask harder questions. Talent may become harder to keep. The race may continue, but the cost of ignoring internal doubt grows.
Coxon’s proposed solution is broad. He asked fellow engineers to reflect on their work. He also called on regulators to create international agreements that govern the pace of development, including a temporary ban on advancing AI’s core capabilities. Whether such a pause is realistic is another question. Still, his post has turned a private unease into a public debate, and the fintech world is watching.
