An alarming rise in AI-driven cyber-attacks has been anticipated for next year.

With threat actors expected to rapidly advance their use of AI to enhance offensive tactics, according to Google Cloud’s Cybersecurity Forecast 2025.

Cybercriminals are set to leverage large language models (LLMs), deepfakes, and AI-powered reconnaissance tools to create more sophisticated social engineering attacks and cyber espionage strategies.

The report warns of the evolving landscape in AI for Information Operations (IO), predicting that generative AI will be increasingly used to produce fake personas and content at scale, amplifying reach and impact.

Further concerns include the rising geopolitical tensions that could drive state-sponsored cyber activities from countries like Russia, China, Iran, and North Korea, targeting critical industries, government sectors, and public services worldwide.

The EMEA region faces unique risks as rapid cloud adoption and increased misconfigurations leave cloud environments vulnerable. Google Cloud’s Mandiant response teams have already seen a spike in investigations involving inadequate cloud monitoring, weak security protocols, and credential reuse, signalling an urgent need for stronger cloud security strategies in 2025.

Andy Ward, SVP International at Absolute Security commented: "With AI being increasingly used in sophisticated cyber-attacks, Google Cloud’s latest forecast emphasises the urgent need for organisations to prioritise cyber resilience. As tactics like social engineering, deepfakes, and advanced reconnaissance evolve, proactive security measures are essential to protect sensitive data from potentially devastating breaches. While AI offers advantages for workforce development, our research shows that 54% of CISOs feel their security teams are unprepared for evolving AI-powered threats. Organisations must implement robust network visibility, control, and a strong cyber resilience framework to defend against emerging AI-driven threats."