Anthropic has introduced a new tool, Claude Code Security, that is stirring considerable disruption in the cybersecurity sector. Launched two weeks after major SaaS stocks in the United States and Israel faced volatility due to AI tools, the release has raised concerns regarding the viability of existing business models within the industry. This latest product from the company behind the Claude chatbot employs its Claude Opus 4.6 model to analyze software code, mimicking the approach of a human security researcher rather than relying solely on traditional rule-based detection.
The tool’s capabilities include tracking data flows within applications, identifying business logic flaws, and conducting multi-step validation, which incorporates AI-driven self-review processes aimed at reducing false positives. While it proposes automatic fixes for developers’ approval, it currently lacks the functionality for runtime testing, meaning it does not provide real-time protection against potential threats.
Anthropic claims to have tested the system on active open-source projects, uncovering over 500 previously unknown vulnerabilities. This development comes after more than a year of efforts involving its Frontier Red Team, cybersecurity competitions like Capture the Flag, and partnerships with research institutions.
Market reactions were immediate. Shares of prominent cybersecurity companies such as CrowdStrike, Okta, Cloudflare, and Zscaler experienced sharp declines following the announcement. In Israel, stocks were similarly affected: JFrog plummeted by 24%, Check Point fell 4%, and SentinelOne and Palo Alto Networks slipped by nearly 3% and 1.5%, respectively.
Investor apprehensions center around the possibility that AI systems capable of autonomously scanning and rectifying code may threaten traditional security analysis tools, potentially squeezing profit margins for companies whose products rely on AI-driven detection methodologies. However, some industry experts caution against overreacting. Liran Grinberg, founding partner at venture capital firm Team8, described the market response as disproportionate, suggesting that many affected firms have limited exposure to the segment that Anthropic is targeting.
Grinberg also emphasized that although the entry of significant AI model developers into the cybersecurity landscape was anticipated, the intricate nature of enterprise-wide security infrastructure demands operational expertise that cannot be replicated swiftly. Kobi Samboursky, a partner at Glilot Capital, echoed this sentiment, asserting that he does not foresee a dramatic downturn in the industry. “The expertise of cybersecurity companies remains critical,” he stated. “Large organizations will not rely solely on a generic AI tool.”
Tomer Perry, CEO of InnoCom Group Aman, noted that recent market trends indicate an almost automatic reaction to every new AI product. He stated, “The battles in cybersecurity remain the same. They are simply becoming more technological.”
Industry analysts acknowledge that junior cybersecurity roles and startups focusing on narrow AI-based solutions may encounter challenges if companies opt to use general AI tools for similar tasks internally. Additionally, the potential for malicious use of such tools raises further concerns. While enhanced detection technology could complicate the work of cybercriminals, these actors might seek to exploit similar AI capabilities for their purposes. Anthropic has indicated that access to its new tool will be limited to mitigate such risks.
Interestingly, comparable products from competitors, including OpenAI’s Aardvark, launched in October 2025, alongside Microsoft’s Security Copilot and Google’s Security Command Center, did not trigger the same level of market disruption as Anthropic’s announcement. Itai Schwartz, co-founder and CTO of cybersecurity firm MIND, noted, “It is not another code-scanning tool that defines enterprise security, but the ability to manage risk end-to-end. AI can identify problems, but it does not replace cybersecurity strategy, organizational accountability, or operational complexity.”
Looking ahead, Anthropic has expressed optimism, stating it anticipates a significant share of the world’s software code will be scanned by AI in the near future. For the cybersecurity industry, this forecast could represent not extinction but a transformative shift that will reshape operational practices and strategies.
See also
Anthropic’s Claims of AI-Driven Cyberattacks Raise Industry Skepticism
Anthropic Reports AI-Driven Cyberattack Linked to Chinese Espionage
Quantum Computing Threatens Current Cryptography, Experts Seek Solutions
Anthropic’s Claude AI exploited in significant cyber-espionage operation
AI Poisoning Attacks Surge 40%: Businesses Face Growing Cybersecurity Risks
