[DigitalToday reporter Chi-gyu Hwang (황치규)] The security industry is stepping up efforts to respond with AI to security threats that abuse AI models. Recently, the industry has moved faster toward agentic AI-based security.

Zscaler introduced two security products for AI agents and an "AI Access Graph." The company said it reflects a situation in which corporate environments are shifting from human users to autonomous AI agents, making it difficult to control the exercise of privileges and data flows with only existing security systems. Check Point launched Agentic Exposure Validation for Exposure Management (AEV) to help defenders respond on equal footing against AI-based attackers.

ㆍZscaler launches AI broker and endpoint AI security, controls agent privileges and communications ㆍStop autonomous AI attacks...Check Point launches agent-based threat exposure validation solution

As AI can find vulnerabilities and write actual attack code within hours, concerns are growing over a so-called "patch gap," in which corporate security teams have less and less time to fix vulnerabilities. Google M-Trends 2026 report said there are real cases in which attacks occur before patches are deployed.

ㆍNo time to respond to ultra-fast AI attacks...'patch gap' emergency

On this, Tanium Korea made clear that for corporate security strategy in the post-MITOS era, patch management should also be reorganised around AI-based automation as AI speeds up cyberattacks.

ㆍSecurity in the MITOS era?..."Policy-based autonomous patching is the answer"

Data security company Rubrik formally launched "Agent Cloud," a security service dedicated to Anthropic Claude Code and Claude Cowork. Rubrik co-founder and CEO Bipul Sinha (비풀 신하) said, "AI agents can be abused for external attacks or cause problems during internal deployment," adding, "Rubrik AI provides resilience against machine-speed cyber intrusions and compromised agents."

ㆍRubrik launches security product dedicated to Claude Code and Cowork, includes 'rollback AI actions'

IT market research firm Gartner identified four major cyber threats that could allow attackers to gain an advantage by exploiting targeted vulnerabilities: deepfakes, AI application compromise, prompt injection and the software supply chain.

ㆍ"AI-driven cyber security threats...need a multilayered defense strategy"

Other domestic and overseas corporate moves and issues around security were also reviewed.

Samsung SDS is strengthening its AI-based cloud security business by cooperating with U.S. AI security startup XBOW and local cloud security company Tatum Security. Digital forensics specialist Uracle is launching "DFAS Pro Mac," a forensics solution for macOS, and expanding its forensics response beyond Windows to Mac OS. Cybersecurity company Stealien is launching "WebSuit," a web defacement prevention and protection solution that applies AI-based behaviour analysis, and expanding from mobile apps into web service protection. JiranJigyo Security subsidiary Subssoft signed a strategic partnership with Japanese security company JSecurity and Exosphere Labs and will expand its integrated mobile and PC security solution business in Japan.

ㆍSamsung SDS expands AI and cloud security business, cooperates with domestic and overseas security specialists ㆍUracle launches DFAS Pro Mac digital forensics solution for macOS, expands to multiple platforms ㆍStealien launches WebSuit, expands from mobile apps to web service security ㆍSubssoft cooperates with JSecurity, enters Japan mobile antivirus market

OpenAI added data loss prevention and login session management features to ChatGPT. TrendAI, the enterprise cybersecurity unit of security company Trend Micro, is participating in Project Glasswing, in which Anthropic discloses its AI model "MITOS" to selected companies specialising in cybersecurity uses. According to foreign media reports, Anthropic is sending engineers to the U.S. National Security Agency (NSA) to support the use of the MITOS AI model for offensive cyber operations. Apple is introducing a feature that automatically changes leaked or vulnerable passwords using artificial intelligence, moving to strengthen account security.

ㆍOpenAI expands security features in ChatGPT, adds data loss prevention and login session management ㆍTrendAI joins Anthropic Glasswing project ㆍAnthropic sends engineers to NSA, supports cyber attack operations with MITOS ㆍApple unveils security features, supports automatic password updates

As the financial sector accelerates its AI transformation (AX), response systems for digital risks such as hacking and voice phishing are emerging as key tasks. Financial authorities called for widening the use of AI in the financial sector through steps including easing network separation rules, while also ordering improvements to internal controls and consumer protection systems to respond to security threats from the spread of high-performance AI and the sophistication of financial fraud.

ㆍIn the AX era, security and consumer protection must also fit...financial sector revamps response systems