Newsbite analysis · AI, Security
The Economic Times
Anthropic opens its most powerful AI models to more security teams
Anthropic has revamped its Cyber Verification Program (CVP) to permit more vetted cybersecurity teams to access its advanced AI models, specifically Claude Opus and Sonnet. This follows the success of Glasswing, which identified over 129,000 vulnerabilities in critical software. The program is designed in tiers to address varying security needs, ensuring extensive checks on participant organizations.
Act nowConfidence: HighShort term horizonStory: First report
EvidenceOfficial announcementNo quotes
Published 6 Oct 2026, 20:30 UTC5 min read
Lead image — The Economic Times
Brief
NewsBite reading:Anthropic expands access to powerful AI models for security teams
TL;DR
- Anthropic's Glasswing initiative identified 129,000 vulnerabilities.
- The revamped CVP now allows security teams to access powerful AI models.
- Three tiers of access cater to different security requirements.
- Glasswing's findings are expected to be an undercount.
- More than 33,000 vulnerabilities rated as critical or high severity.
Interesting facts
- The tiered program includes a specialized group for critical systems like power grids.
- The new program is expected to highlight AI's potential risks and benefits to software security.
What changed
The expansion of the Cyber Verification Program allows more security teams to utilize Anthropic's AI models with fewer safeguards.
Unchanged: The commitment to rigorous vetting of organizations accessing these models remains intact.
Decision signal
Decision brief
Actionable takeaways
- Cybersecurity teams should consider applying for the new CVP tier access.
- Firms should audit their software for vulnerabilities using advanced AI tools.
- Remain vigilant of the evolving landscape of AI capabilities in security.
- Prepare for potential increased scrutiny and security measures from regulators.
Decision signal
Act now
Action required
None
Event status
Announced
Technical significance
Meaningful
Competitive impact
Strengthens position
Source reliability
High
Impact certainty
High
Novelty
High
Execution risk
Medium
Strategic weight
High
Trend: Accelerating
- AI in cybersecurity
- Vulnerability management technology
Sentiment signal
Sentiment
Overall: PositiveMood: Bullish
The news presents an optimistic outlook on improving cybersecurity through advanced AI tools, despite inherent risks.
Drivers
Increased access to powerful AISignificant vulnerability detectionEnhanced collaboration with security professionals
By category
AIPositive
This initiative enhances the application of AI in real-world security challenges.
SecurityPositive
The increased access to AI tools will aid in managing and overcoming security vulnerabilities.
Notable entities
AnthropiccompanyPositive
Anthropic is advancing its cybersecurity initiatives through AI, enhancing its reputation in the field.
Market signal
Impact & reach
Anthropic's opening of AI models to more security teams is significant for enhancing software security.
The expansion of access to advanced AI tools could lead to a significant reduction in exploitable vulnerabilities in critical software systems, enhancing overall cybersecurity. This move also raises concerns about the potential misuse of AI in software hacking.
security teamsPositive
Security teams will have enhanced tools to identify and mitigate software vulnerabilities.
Geographic focus
USPositive
The collaboration includes involvement from the US government in vetting participants.
Likely winners
cybersecurity firmsAI development companies
Risk signal
Risk radar
Cybersecurity riskMedium
The potential for misuse of AI in hacking remains a concern.
Data governance riskMedium
Increased access to AI may present new privacy and governance challenges.
Reputational riskLow
Successful implementation could enhance public perception of Anthropic.
Execution riskMedium
The successful deployment of the program will depend on rigorous vetting processes.
Infrastructure riskLow
The initiative supports safeguarding critical infrastructure.
Geopolitical riskMedium
The program involves US government participation, which could affect geopolitical stances on technology.
Regulatory riskMedium
The use of AI in cybersecurity may prompt regulatory scrutiny and compliance requirements.
Supply chain riskLow
As it involves software security tools, it may pose limited direct supply chain risks.
Talent displacement riskLow
There may be new job opportunities within the cybersecurity sector.
AI liability riskMedium
As AI tools become more powerful, liability for misuse could rise.
Sourcing
Sources & evidence
The automated analysis found no sources named in the text.