Surprisingly Useful AI Article Enhancements
Artificial intelligence (AI) has reshaped the cybersecurity landscape, putting businesses at risk of increasingly sophisticated attacks. The ability of AI to pattern-match multiple sources and generate realistic deep fake materials has made social engineering more complex and challenging to identify. AI-powered automated tools have also made it easier for bad actors to implement large-scale attacks at unprecedented speeds.
According to Veeam’s 2024 Data Protection Report, over eight in 10 organisations in Asia Pacific and Japan (APJ) suffered at least one attack last year. Additionally, only six out of 10 servers impacted could recover within an acceptable timeframe after a crisis. Businesses urgently need to adopt strategies in protecting against evolving threats. Preparation for recovery in the unfortunate event of an attack is a must, not a nice to have.
Basic cyber hygiene continues to be key
While the complexity of AI may seem daunting, defending against AI threats is a similar approach to employing strategies in protecting against any cybersecurity threat. Fundamental cyber hygiene practices, including multi-factor authentication (MFA), phishing training and patch management, remain critical to protect against AI threats. While it is challenging, if not impossible, for businesses to mitigate all cyber risks, it’s important to steer away from an all-or-nothing approach. The most basic cyber best practices can go a long way in reducing the risk and impact of an attack.
Leveraging counter AI strategies
As more businesses begin leveraging AI, they must simultaneously use AI to improve threat detection and visibility.
For example, algorithms that use AI to analyse network traffic and detect unusual behaviour can help uncover malicious activity promptly and initiate preventative measures. AI can also enhance patch management by identifying high-risk vulnerabilities and prioritising remediation. Further, AI-powered analytics can provide valuable insights into emerging threats and prompt businesses to strengthen specific areas of their security posture proactively. Veeam’s partnership with Microsoft is focused on bringing this to customers, by way of integrating Microsoft Copilot for automated data analysis, making it easier and more cost-effective to visualise data and generate actionable insights.
As AI technology advances and becomes more accessible, we can expect to see a shift towards privately owned large language models (LLMs). Private LLMs allow control over inputs and outputs, ensuring sensitive information is kept within the corporate network, outputs incorporate the appropriate technical jargon, and the model aligns with local regulations. Algorithms can also identify anomalies more precisely when the LLM is tailored to the organisation’s unique data landscape.
Data resiliency will play a critical role in the AI landscape
In the face of rapidly evolving threats, businesses must also have a robust backup and recovery strategy that ensures data resilience. Resiliency enables organisations to maintain absolute control when encountering AI-driven manipulation or damage, providing a safety net and protection against extended downtime following an attack.
Concerningly, Veeam’s 2023 Ransomware Trends Report found that in almost all cyber events, bad actors attempted to attack backup repositories, meaning backups must be present and survive attacks. This is why Veeam endorses the 3-2-1-1-0 data backup rule, which recommends maintaining at least three copies of its data across two different media, with at least one copy offside and at least one copy offline, air-gapped or immutable. The backups should also be verified without errors, which means they are readable and reliability recoverable.
Businesses should also consider strategies to contain the spread of malicious activity and minimise the impact of attacks. For example, lateral movement – the act of gaining access to other areas of a network after obtaining initial access – is a common attack technique that AI can expedite. Data segmentation policies that divide networks into isolated segments can limit this type of movement, thereby reducing attack surface and impact.
AI will continue to reshape the cybersecurity landscape, making cyber attacks a when rather than an if. Businesses must not only strengthen preventative strategies, but also ensure they can recover swiftly in the event of an attack. Fortunately, fundamental cyber best practices, a robust backup and recovery strategy, and AI-powered tools can significantly enhance cyber resiliency, safeguarding businesses against AI threats.
Dave Russell is Vice President of Enterprise Strategy, Veeam.
Last Updated on July 29, 2024 by Dave Russell
