What happened
Anthropic employee Evan Hubinger, who previously worked at Ripple, has drawn attention following an assessment of extreme AI risk.
Why it matters
The statement has drawn attention to assessments of catastrophic AI risks while simultaneously highlighting the author's connection to the crypto industry, though its context remains undisclosed.
Anthropic employee Evan Hubinger has attracted widespread attention after stating there is a 10% probability that artificial intelligence could wipe out humanity. According to a Protos report, he previously worked at Ripple.
The news is significant because the public discussion of existential AI risks involves not only the research community but also an individual with experience in the crypto industry. However, the source does not provide details regarding Hubinger's role at Ripple or the circumstances surrounding his statement.
The available data is presented as a brief Protos synopsis rather than a full text or independent confirmation. Therefore, it is currently impossible to establish the context in which the 10% estimate was made or how it should be interpreted.
Confirmed facts
- Evan Hubinger works at Anthropic.
- Evan Hubinger previously worked at Ripple.
- Hubinger gained widespread attention after stating there is a 10% probability that artificial intelligence will wipe out humanity.
- Сведения опубликованы Protos 10 сентября 2026 года.
- The available confirmation is presented as a synopsis of source metadata rather than the full text of the publication.
Context
The material is based on a single independent source with limited access to content: the package contains only a synopsis of Protos metadata.
What remains unknown
- In what context did Evan Hubinger cite the 10%? probability?
- What arguments did he present to support this assessment?
- What position did Hubinger hold at Ripple?
- Are there any comments from Anthropic or Ripple regarding this matter?
- Does the assessment reflect Hubinger's personal opinion or the stance of an organization?
Editorial context
Confidence: medium
The likely consequence is further discussion on methods for assessing catastrophic AI risks, rather than an immediate change in company policies. The next observable signal may be the full publication with context for the statement or comments from Anthropic and Ripple. Significant uncertainty remains due to the lack of a full text and independent confirmation.