Microsoft Warns AI is Supercharging Cyberattacks as Hackers Automate Phishing, Malware & Recon

Artificial Intelligence (AI) is rapidly reshaping the cybersecurity landscape, and Microsoft now says the same technology helping businesses move faster may also be giving cybercriminals new tools. In a recent report, Microsoft Threat Intelligence warned that threat actors are increasingly integrating AI into their attack workflows, allowing them to scale operations and automate tasks that once required specialized skills.

At a time when companies are racing to deploy AI across productivity tools, coding platforms, and enterprise systems, attackers appear to be experimenting just as aggressively. The report suggests that AI is gradually becoming a “force multiplier” in cybercrime, helping adversaries streamline everything from reconnaissance to social engineering.

Microsoft warns AI is accelerating modern cyberattacks

According to Microsoft Threat Intelligence, malicious groups are now using AI tools to reduce friction across multiple stages of a cyberattack. These include early reconnaissance, phishing campaigns, malware development, and post-compromise activity once systems have already been breached.

Generative AI tools, particularly large language models, are reportedly among the most commonly misused technologies. Attackers can use them to craft convincing phishing emails, translate stolen information, summarize large volumes of data, or even assist with writing and debugging malicious code.

Microsoft notes that AI is not fully replacing human operators. Instead, threat actors appear to be using it to speed up tasks that would otherwise take hours or days. The result, the company suggests, is a faster and more scalable attack cycle. The report also highlights activity linked to several state-backed groups. Some attackers have reportedly used AI to generate fake identities, write job applications, and maintain long-term access inside targeted organizations.

Another emerging concern involves attempts to bypass AI safety controls. Attackers are experimenting with so-called “jailbreak” techniques that manipulate prompts to make AI systems generate restricted or harmful outputs.

The next phase of AI-powered cyber threats may already be forming

Looking ahead, Microsoft says more advanced AI-powered threats are beginning to surface. Early experiments reportedly include malware capable of generating scripts dynamically or adapting behavior during execution rather than relying entirely on prewritten instructions.

Researchers are also watching the development of agent-based AI systems that can plan multi-step tasks, evaluate results, and adjust actions with limited human oversight. For now, many of these capabilities remain experimental. However, as AI tools become more powerful and accessible, it appears increasingly likely that cybercriminals will continue refining how they use them.

via: Beeping Computer