Verizon becomes first telco in Anthropic Project Glasswing

TL;DR: Verizon joined Anthropic's Project Glasswing as the first telecommunications company in the initiative. Anthropic introduced its Claude Mythos model to a limited group of companies on 7 April, including Amazon Web Services, Apple, Nvidia and Google. Anthropic said it committed up to $100 million in model usage credits and $4 million in donations to open-source security organisations.

Verizon joined Anthropic's Project Glasswing as the first telecommunications company in the coalition. Anthropic said the initiative is focused on securing critical infrastructure for the AI era.

Anthropic introduced its Claude Mythos model on 7 April through Project Glasswing to a limited number of technology companies rather than releasing it publicly. The companies named were Amazon Web Services, Apple, Nvidia and Google. Anthropic said Mythos Preview can autonomously identify and exploit software vulnerabilities, and said it restricted access because of both the model's defensive potential and the risks of broader availability.

Anthropic said it committed up to $100 million in model usage credits to support Project Glasswing and $4 million in donations to open-source security organisations. Verizon said its information security team spent several months evaluating Mythos Preview, and said all testing was conducted under strict usage policies and safety standards. Verizon CEO Dan Schulman said the operator is the only telecommunications company using Mythos Preview and will share cross-industry insights from the work.

Related questions

• What is Anthropic's Project Glasswing?
• When did Anthropic introduce Claude Mythos?
• Which companies were given access to Mythos Preview?
• How much funding did Anthropic commit to Project Glasswing?
• Why did Anthropic restrict access to Mythos Preview?