AI in Cybersecurity 2025:
Harnessing Good AI to Combat Evolving Cyber Threats

Awareness

Cybersecurity 101

Published on 26.02.2025

4 min read

Get in Touch

Google Cloud’s Cybersecurity Forecast 2025 highlights the dramatic and continuous development as well as scaling of social engineering attacks, with cybercriminals leveraging the newest AI advancements.

The good news: AI is also rapidly bringing new tools for threat detection and response. Sunil Potti predicts that 2025 will be the year “AI becomes a key tool in combating threats” (Google Cloud Security, 2025). As both attackers and defenders leverage AI, the challenge will be focusing on leveraging "good" AI to mitigate "bad" AI.

The Evolving Threat Landscape

The rise of AI is making hackers smarter and more dangerous. Over the past year, we have seen a surge in sophisticated phishing attacks, which exploit generative AI models to craft convincing messages that can evade spam filters. Cybercriminals are now able to:

Automate and personalize attacks: AI tools enable the rapid generation of phishing campaigns tailored to specific individuals, making them harder to detect.
Leverage GenAI for adaptability and creativity: Attackers can continuously tweak their strategies, creating a dynamic threat landscape.
Bypass defenses with localization: AI-driven tools improve the localization and customization of attack content, circumventing traditional spam filters.

AI-Powered Criminal Tools

Deepfake audio and video technology have also reached new levels of realism. In 2024, multiple organizations fell victim to AI-enhanced scams:

Ferrari narrowly escaped a deepfake scam when a manager identified inconsistencies in the voice of a supposed CEO and asked a personal question which the AI could not answer.
Arup, a British firm, lost $25 million in a deepfake scheme.
Retailer Pepco suffered a €15 million loss due to a phishing attack.

The darknet is flooded with "evil" AI tools like FraudGPT and WormGPT, explicitely designed for malicious purposes. These tools, ranging from $200 to $1,700, lower the barrier to entry for cybercrime, empowering even amateur attackers to launch sophisticated campaigns.

“AI becomes a key tool in combating threats”

— Sunil Potti, VP/GM of Google Cloud Security

AI as a Cyber Guardian for Organizations

Despite the “bad”, AI is also a “good” ally in the fight against cybercrime. Defenders can harness the same tools used by hackers to create effective phishing training campaigns across multiple channels, including deepfake simulations. With 9 out of 10 cyber attacks involving human error, organizations must build strong "human firewalls" through comprehensive training programs.

According to Google Cloud’s Cybersecurity Forecast 2025, organizations must proactively focus on “staying ahead of emerging threats through continuous monitoring and threat intelligence” (Potti, 2024). Additionally, as regulations like NIS2 and GDPR evolve, CISOs must adapt their security measures to ensure compliance.

Companies like Company Shield are already leveraging AI to prepare employees to recognize and respond to increasingly sophisticated attacks AI is a game changer in cybersecurity, for both good and bad, and the future of cybersecurity depends on how effectively we use AI for good.

Sign up for a free demo with a Company Shield team member or try our Instant Demo right away.