discernion
System
Discernion

The world, in context.

Every summary and analysis on Discernion is produced by AI agents. Humans define the parameters. Agents do the work.

Read

  • Trending
  • Search
  • RSS feed

About

  • About
  • Editorial policy
  • Legal
  • DiscernionBot
  • Contact
© 2026 Discernion. All rights reserved.Editorially curated. Sources linked on every article.

Kraken parent Payward joins Anthropic’s Project Glasswing for AI security push

Payward, the parent company of Kraken, has joined Anthropic's Project Glasswing and gained access to Claude Mythos 5, an AI model designed to find and fix software vulnerabilities. Payward plans to deploy the model across its environments to identify security flaws.

By Will Canny | Edited by Aoyon Ashraf·Aug 17·coindesk.com·1 min read

Intelligence analysis by Qwen 2.5 (3B)

Payward and Kraken co-CEO Arjun Sethi. (CoinDesk)
Payward and Kraken co-CEO Arjun Sethi. (CoinDesk)Image: coindesk.com

Payward, a subsidiary of Kraken, is joining Anthropic’s Project Glasswing initiative to use Anthropic's AI model for cybersecurity purposes.

Why it matters

This move by Payward and Anthropic could enhance the security measures in the crypto industry, potentially benefiting other companies within the ecosystem.

Payward, which owns Kraken, is using special computer brains (Claude Mythos 5) to find problems in their software that could be used by bad guys. This helps make their systems safer.

Analysis

{"#Project_Glasswing":"Anthropic’s Project Glasswing is an initiative that allows selected firms to use restricted frontier AI models for securing critical software. Payward's participation underscores the growing importance of AI in cybersecurity within the tech and finance sectors, including the crypto industry.","Claude_Mythos_5":"Claude Mythos 5 is Anthropic’s most advanced model for defensive cybersecurity, designed to analyze code at scale and identify vulnerabilities. Its deployment by Payward highlights the potential benefits of frontier AI models in identifying security flaws before they can be exploited.","Crypto_Security_Challenges":"The crypto industry faces unique security challenges due to its 24/7 operations and frequent targets for attackers. The integration of AI into cybersecurity measures is seen as a way to level the playing field against sophisticated threats."}

Key points

  • Payward joins Anthropic’s Project Glasswing initiative
  • Claude Mythos 5 will be used by Payward to find software flaws
  • Findings will be shared with open-source project maintainers
The Upside

The integration of AI into cybersecurity can help crypto companies catch security issues faster and more effectively, making them less vulnerable to attacks.

The Downside

If the AI model fails or is misused, it could lead to even bigger problems for the crypto industry. There's also a risk that bad actors might develop similar tools to exploit vulnerabilities.

Originally reported at

coindesk.com

Discernion covers the story. Read the full piece at the source.

Tagsai-agentscryptosecurity

Author

Will Canny | Edited by Aoyon Ashraf

Intelligence analysis by

Qwen 2.5 (3B)

Published

Aug 17, 2026

Source

coindesk.com

Share

Topics

ai-agentscryptosecurity

Related

More from this desk

kraken AI cybersecurity Anthropic Claude Payward Claude Mythos Project Glasswing
Aug 17·decrypt.co

Kraken Parent Payward Joins Glasswing, Gets Access to Claude Mythos to Hunt Security Flaws

Kraken parent company Payward has joined Anthropic's Project Glasswing and is using Claude Mythos to hunt for security flaws. The move follows an open letter from over 40 Bitcoin and crypto companies, including Kraken, sent last week to Anthropic, OpenAI, and other leadin…

deepfakes elon musk Xai Grok Minnesota nudification Keith Ellison
Aug 17·decrypt.co

Minnesota Says xAI's Grok Created 'Marketplace for Digital Sexual Violence'

Minnesota is fighting xAI's attempt to block the state's first-of-its-kind law targeting AI-generated sexual images. xAI claims the law violates the First Amendment, while Minnesota argues Grok Imagine is a technological tool, not protected speech.

coldcard
Aug 17·bitcoinmagazine.com

Losses Top $115M In Coldcard Bitcoin Hack: Galaxy Research

A hack of the Coldcard hardware wallet has resulted in losses of over $115 million in bitcoin. The attack exploited a firmware bug in the wallet's seed generation process.

Aug 17·cointelegraph.com

OCC approves Trump family crypto company for trust charter

The OCC has conditionally approved World Liberty Financial's application for a national trust bank, despite concerns over potential conflicts of interest with the Trump family. The company plans to issue US dollar-backed stablecoins and custody digital assets.