In the AI Era, Hacking Defense Is a Corporate Survival Issue

Even after recovering data by paying ransom after a ransomware infection, there is a high risk of reinfection if the path the hacker used to infiltrate is not blocked. As seen in the case of a famous online bookstore, the risk persists if temporary recovery is the only measure taken without fundamental security actions. Hackers tend to reuse paths they have already breached, so continuous management is critical.

Management must first identify what the most important data within the company is. It is essential to know the exact location of critical data such as source code, blueprints, and transaction information, and to minimize access privileges. Unnecessary access rights, such as those for departed employees, must be immediately deleted, and thorough management of file sharing with partners and maintenance staff accounts must also be implemented. Care should be taken not to set simple passwords.

Many companies tend to perceive security risks as not immediately urgent. However, a security incident is like a fire; it can spread damage not only to the company itself but also to customers and partners. It is crucial to take preventive measures before the damage spreads extensively.

With the advancement of AI technology, the hacking business has become corporatized and systematically compartmentalized. Hacking, which once required highly skilled experts, can now be carried out extensively by low-skilled criminals thanks to AI. Hacking has evolved into a corporate-like structure, from target scouting, infiltration, negotiation, to even after-sales support (AS).

As AI models quickly identify security vulnerabilities on their own, the speed at which hackers can find entry points has dramatically increased. Comparing Windows security vulnerability updates, we can see a significant increase of about 2 to 5 times. This leads to a situation where the speed of defense cannot keep up with the speed of attack.

Employees unauthorizedly inputting company documents into AI serves as a trigger for confidential information leaks and expanded security vulnerabilities. If a company completely bans AI use, employees may resort to 'shadow IT,' secretly using personal accounts or devices. This risks making information leakage routes uncontrollable, outside the company's management scope.

Clear internal information input standards must be established to prevent the leakage of core assets such as customer information and contracts. When using generative AI, administrators need a procedure to directly verify data storage methods and server locations. Furthermore, it is important to implement security solutions to build a system that proactively blocks the leakage of sensitive information.

Violation of personal information protection laws can result in fines of up to 10% of sales. For example, a company with 5 billion won in sales might have to pay 500 million won in fines, which is a significant loss that is very difficult to cover with operating profit. Beyond direct fines, additional damages occur from incident investigation, lawyer fees, and customer attrition, making incident recovery costs far greater than security investment costs.

While the CEO does not need to set up firewalls directly, they must always ensure that the company's critical information is stored, well-protected, and accessible only to authorized personnel. It is crucial to devise strategies to minimize damage and quickly restore operations in the event of an incident.

Government support measures are continuously being introduced to help SMEs effectively strengthen their security. Examples include the Korea Internet & Security Agency (KISA)'s free vulnerability assessment program and local government security support programs. Efforts are needed to find and actively utilize such information.

Such support helps with incident recovery and enables rapid dissemination of cases to help similar industries prevent similar damages.

Attackers attempt attacks using sophisticated themes that employees are highly likely to click on. Simulated hacking drills are crucial to enhance employees' response capabilities by creating realistic training emails using generative AI. This helps employees develop the ability to recognize threats and not fall for attacks.

In the past, hackers primarily targeted large corporations, but now AI scans systems and randomly infiltrates companies with security vulnerabilities. This means even small businesses can easily be exposed to automated AI-based attacks. Therefore, a CEO's regular security interest becomes a critical defense shield for protecting the company.