ハッカーがAnthropicの「危険すぎて公開できない」AIモデルMythosに侵入
ハッカーがAnthropicの安全対策を突破し、「危険すぎて公開できない」とされていたAIモデル「Mythos」に不正アクセスしたとの報道。このモデルは内部テスト用に開発されたもので、一般公開はされていなかったが、外部からの攻撃によりその存在と一部の機能が暴露された可能性がある。セキュリティ専門家は、高度なAIモデルを保護するためのさらなる対策の必要性を指摘している。
関連記事
Amazon allegedly coordinated with brands like Levi's and Hanes to pressure competitors such as Walmart and Target to raise prices. When brands sold products cheaper on other sites, Amazon would suppress their listings, leading brands to demand price increases from competitors. Multiple antitrust cases are now scheduled for 2027 regarding these alleged price manipulation practices.
A small group of unauthorized users in a Discord channel gained access to Anthropic's Mythos AI model on the same day it was announced for limited testing, and have been using it regularly since, according to Bloomberg. The group has also accessed other unreleased Anthropic AI models. Anthropic had described Mythos as powerful enough to enable dangerous cyberattacks.
The NHS is reportedly preparing to close nearly all of its Open Source repositories, a move that has disappointed a former government digital worker who championed Open Source during his time in roles at GDS, NHSX, and i.AI. The author criticizes NHS England for backtracking on its commitment to Open Source.
Anthropic's Mythos research preview reveals insights about frontier AI models, sandbox escapes, and emerging cybersecurity risks. The analysis examines how these developments may impact internet security frameworks.
A report claims Anthropic has surpassed OpenAI in annual revenue. If confirmed, this represents a significant shift in the competitive landscape between the two AI companies.