Security companies are actively moving to strengthen AI security while expanding their businesses beyond traditional security domains.

As more companies that did not previously focus on security expand into it as part of their AI governance strategies, security firms are increasingly moving beyond security into AI governance and control as well.

PRIBIT AI, as part of efforts to shift its business identity from a security company to an AI governance and risk-control firm, changed its name from Private Technology and also announced an execution strategy for the transition. Global company F5 is also expanding beyond security into AI governance and cost management. F5's recently introduced AI gateway applies consistent policies to all AI requests and provides an integrated control plane that manages how AI models, agents and tools are accessed and used. This can also improve cost efficiency in large-scale AI operations, the company said.

• PRIBIT AI reveals strategy to transform from security to an AI governance and risk-control company • [D2People] "Enterprise AI is hard to expand without security"

Among South Korean companies, some are joining Daybreak, an AI-based cybersecurity cooperation chain led by OpenAI. Samsung SDS is participating in the Daybreak Partner Program as a pilot partner and is set to expand its AI-based cybersecurity business in earnest.

• Samsung SDS to push ahead with AI-based security business, joining OpenAI's Daybreak • S2W joins OpenAI's global cybersecurity initiative Daybreak

Other moves and issues involving companies at home and abroad around security were also compiled.

MonitorLab, a B2B SECaaS (Security as a Service) specialist, is speeding up its push into the generative AI security market. MonitorLab recently advanced its GenAI security functions to support granular security policies tailored to department and job characteristics, and to expand the scope of controls to developer environments. Aton signed a contract with Eastar Jet to supply its PQC-based mobile OTP solution Quantum SafeOTP for authentication of employee access rights to internal work systems. MegazoneCloud obtained the AWS Security Competency from Amazon Web Services (AWS). S2W launched DRI (Deep Research Investigator), an autonomous AI agent specialized for criminal investigations. Enki WhiteHat, an offensive security specialist, supplied KB Kookmin Bank with OFFen AI DAST, an AI-based automated web application penetration-testing solution.

• MonitorLab to hold IASF2026, launching a major push into the generative AI security market • Aton supplies post-quantum cryptography solution to Eastar Jet, expanding clients beyond finance • MegazoneCloud wins AWS Security Competency, expanding AI security • S2W launches DRI, autonomous AI agent specialized for criminal investigations • Enki WhiteHat supplies KB Kookmin Bank with AI-based automated web app penetration-testing solution

Cloudflare has begun providing early access to its Vulnerability Detection and Response service, which finds vulnerabilities in customer application software and blocks attacks at the network edge. OpenAI warned that open-weight AI models have relatively low guardrails, and that with fine-tuning and additional computing they can approach frontier-model cyberattack capabilities, meaning companies must guard against misuse.

• Cloudflare detects and blocks code vulnerabilities with OpenAI cybersecurity model • OpenAI: "In the AI era, enterprise security must shift to an agent-centric model"

As AI hacking spreads, companies are not only increasing cybersecurity budgets but are also shifting spending priorities. In this trend, some security technologies benefit, while some existing security products appear to be losing priority. Forecasts that victims of hacking will be AI agents rather than people are also drawing attention. Bugcrowd CEO Dave Gerry said, "Going forward, we will see agents becoming the victims. We will see more cases where agents, not people, get hacked." There was also a case in which cyber attackers used a multi-agent AI framework to steal thousands of credentials in less than 6 hours. As AI-based cyberattacks continue, chief information security officers (CISOs) are emerging as key players not only in hands-on security work but also in executive decision-making.