According to @AnthropicAI on X, Anthropic introduced Project Glasswing, an urgent initiative to secure critical software using its newest frontier model, Claude Mythos Preview, which it claims can find software vulnerabilities better than all but the most skilled humans. As reported by Anthropic’s official announcement on X, the program targets high-impact codebases where rapid, automated vulnerability discovery can reduce risk and remediation time. According to Anthropic’s post, the Claude Mythos Preview model is positioned for offensive security analysis tasks such as code review, exploit pattern detection, and triage support, indicating near-expert performance on vulnerability discovery. For security buyers and dev teams, this implies faster secure SDLC integrations, earlier defect detection, and potential cost savings across penetration testing cycles, according to Anthropic’s stated capabilities on X.

According to @claudeai on X, Anthropic introduced Claude Code Security in a limited research preview that scans codebases for vulnerabilities and proposes targeted software patches for human review, aiming to catch issues traditional tools miss. As reported by Anthropic via the linked announcement page, the tool is positioned to augment secure SDLC workflows by prioritizing exploitable findings and suggesting remediation diffs, which can shorten mean time to remediation for engineering teams. According to the same source, the early access focuses on accuracy with human-in-the-loop validation, indicating near-term use cases in secure code reviews, backlog triage, and compliance readiness for enterprises integrating AI-assisted application security.