36Kr (direct RSS)Apr 16, 2026, 20:12 UTC+8

Anthropic plans to release its Mythos model to UK banking institutions next week

Anthropico计划下周向英国银行业发布其Mythos模型

Anthropic PBC plans to grant UK financial institutions early access to its Mythos model within the next week. The mechanism is the “Glass Wing” program for selected institutions; Anthropic says the model can identify and potentially exploit cybersecurity flaws, while the post does not disclose specs, pricing, or customer count. The key signal is controlled access, not a broad launch.

Read the original ↗Export Markdown

Take

Anthropic is putting Mythos into UK banking as a whitelist exercise, not a normal product launch. The article gives only three hard hooks: next week, UK financial institutions, and the Glass Wing program. Specs, pricing, and the number of institutions are not disclosed. The loud part is Anthropic’s own framing: Mythos can identify and potentially exploit cybersecurity flaws. That crosses a different line from another coding assistant.

Banks already use models for risk, support, and compliance work. Letting a model touch exploit paths creates a different liability chain. OpenAI and Google have spent the last year selling enterprise AI around audit logs, permissions, and data controls; Anthropic is leading here with staged access because it knows Mythos cannot ship like a regular Claude API. Pip White says she has been close to UK CEOs for a week. I want the regulator, insurer, and incident-liability language.

Source: 36Kr (direct RSS) · 36kr.com
Published: Apr 16, 2026, 20:12 UTC+8 · Apr 16, 2026