Battling the Rising Tide of AI-Driven Cyberattacks

Cybersecurity developer Imperva warns that deficient data protection controls and new generative AI tools based on Large Language Models (LLMs) will trigger an increase in cyberattacks in the next year. Imperva also reveals that insider threats are responsible for 58% of all data breaches and tend to be the most damaging for organizations. 

The integration of AI into the operating systems of companies and governments seems nearly unavoidable. "Prohibiting employees from using generative AI is futile. People inevitably find ways to bypass these restrictions, creating an endless cat-and-mouse game for security teams, without keeping the enterprise meaningfully safer,” says Terry Ray, Data Security GTM and Field CTO, Imperva. While AI tools, such as ChatGPT, have been used to counter hacking attempts, recent studies have revealed that these tools can also serve as a gateway for malicious software.

According to Imperva’s results, 82% of organizations have no insider risk management strategy, hindering their ability to identify employees who may inadvertently leak sensitive information to chatbots. However, carrying out regular security training can enhance employee cybersecurity awareness and aid in identifying unauthorized access granters within the organization.

Furthermore, Imperva´s research shows that 24% of data breaches in the last five years were attributed to human error, enabling malicious use of credentials, theft and potential ransomware attacks. However, insider threats are consistently underestimated by businesses, with 33% stating they do not perceive them as a significant threat. “People do not need to have malicious intent to cause a data breach. Most of the time, they are just trying to be more efficient in doing their jobs,” adds Ray. 

To address these challenges, Imperva recommends shifting the focus from relying solely on employee awareness to a more robust approach centered around safeguarding sensitive data. To accomplish this, Imperva recommends practicing constant visibility into every data repository in their system to ensure that databases are not being misused. 

After creating a data inventory, it is important to classify data assets based on type, sensitivity and value for the organization. By effectively classifying sensitive information, different organizations can prioritize the value of their data, assess potential cyberthreats and determine appropriate mitigation measures.

Ultimately, organizations targeted by hackers should consider carrying out data supervision in a periodic manner. According to Imperva, implementing data monitoring can enhance the organizations' capability of detecting cyberthreats such as anomalous behaviors, data exfiltration and privilege escalation, effectively preventing potential breaches of information.

As AI-powered cyberattacks continue to escalate, businesses and enterprises must prioritize data protection and remain vigilant against insider threats. Embracing comprehensive cybersecurity measures and a proactive approach to safeguarding sensitive information will be paramount in defending in this evolving digital landscape.