Artificial intelligence is transforming industries at an unprecedented pace, but cybersecurity experts are warning that the same technology driving innovation is also empowering cybercriminals. According to a recent report, AI-powered hacking has evolved from a developing concern into a large-scale global threat within just a few months.
Technology companies, governments, and cybersecurity researchers are now facing a new reality where advanced AI models can help hackers automate attacks, identify software vulnerabilities, and scale cybercrime operations faster than ever before. As AI tools become more sophisticated, the line between beneficial automation and dangerous exploitation continues to blur.
Google’s threat intelligence team recently revealed that criminal organizations and state-linked hacking groups are increasingly using commercial AI systems to strengthen cyberattacks. These groups reportedly use advanced large language models such as Gemini, Claude, and OpenAI tools to refine malicious operations and improve attack efficiency.
According to John Hultquist, chief analyst at Google’s threat intelligence group, many people mistakenly believe the AI cybersecurity race is still in the future. However, the reality is that the race has already begun.
Hackers are now using AI to:
This shift marks a major turning point in cybersecurity. Traditional cybercrime often required significant technical expertise and manual effort. AI tools now reduce those barriers, allowing cybercriminals to launch sophisticated attacks with greater speed and efficiency.
Modern AI models are exceptionally skilled at coding and data analysis. While these capabilities help developers and businesses improve productivity, they also create opportunities for malicious actors.
Cybercriminals are reportedly using AI systems to:
Zero-day vulnerabilities are software flaws unknown to developers and security teams. These vulnerabilities are highly valuable because attackers can exploit them before patches become available.
AI systems can rapidly scan and analyze massive codebases, helping hackers identify weaknesses much faster than traditional methods. This significantly increases the risk of large-scale cyberattacks targeting operating systems, browsers, and enterprise software.
AI-generated emails, messages, and fake websites are becoming increasingly convincing. Attackers can now personalize phishing campaigns at scale, making fraudulent communications appear legitimate.
Generative AI can imitate writing styles, corporate communication patterns, and even customer service interactions. This increases the chances of users falling victim to scams, credential theft, and ransomware attacks.
AI tools can assist attackers in writing malicious code, debugging scripts, and improving malware performance. Even less experienced cybercriminals can use AI to create sophisticated attack tools.
Researchers warn that AI-assisted malware development could dramatically increase the volume of cyber threats in the coming years.
Google’s report highlights how quickly AI-powered hacking has escalated. Within only three months, cybersecurity experts observed AI moving from experimental use into industrial-scale cybercrime operations.
The report also claims that some hacking groups are already preparing mass exploitation campaigns using AI-discovered vulnerabilities. This raises concerns about coordinated attacks on governments, financial institutions, healthcare systems, and critical infrastructure.
The rapid adoption of AI in cybercrime demonstrates how technological advances can spread faster than defensive measures.
Another major development involves Anthropic and its unreleased AI model called Mythos. The company reportedly decided not to release the model publicly after discovering it possessed extremely powerful cybersecurity capabilities.
Anthropic claimed that Mythos identified zero-day vulnerabilities in major operating systems and web browsers. According to the company, these discoveries posed potential risks to governments, financial institutions, and global cybersecurity infrastructure.
The company stated that the risks were serious enough to require coordinated defensive action across the technology industry.
This decision reflects growing concerns among AI developers about balancing innovation with public safety.
The report also mentioned experimentation with OpenClaw, an AI tool that gained widespread attention for offering highly autonomous AI agent capabilities with minimal safety restrictions.
OpenClaw became controversial because of reports that the system could perform destructive automated actions, including accidentally deleting large amounts of user data and emails.
Security researchers fear that unrestricted AI agents could become powerful tools for malicious cyber operations if placed in the wrong hands.
Despite growing concerns, not all cybersecurity experts believe AI will completely overwhelm digital defenses.
Steven Murdoch, a professor of security engineering at University College London, suggested that AI could benefit both attackers and defenders. According to Murdoch, cybersecurity is entering a new phase where vulnerability discovery becomes increasingly AI-assisted on all sides.
This means organizations can also use AI for:
AI-driven defensive systems may help organizations respond to threats faster than human teams alone.
However, experts acknowledge that the cybersecurity landscape is becoming far more complex.
Beyond cybersecurity, the article also discussed broader concerns about AI productivity claims.
The Ada Lovelace Institute recently questioned assumptions surrounding large-scale economic benefits from AI adoption in the public sector.
The UK government previously estimated that AI and digital tools could generate £45 billion in productivity savings. However, researchers caution that many studies focus heavily on theoretical efficiency gains without measuring long-term outcomes such as service quality, employee wellbeing, or real-world implementation challenges.
The institute warned that many AI productivity projections rely on untested assumptions and limited evidence.
This highlights a broader issue surrounding artificial intelligence. While AI offers enormous potential, its long-term impact on economies, workplaces, and cybersecurity remains uncertain.
The rise of AI-powered hacking represents one of the biggest cybersecurity challenges of the modern digital era. Businesses, governments, and individuals are increasingly dependent on connected systems, cloud infrastructure, and online services.
As cybercriminals adopt AI technologies, attacks are becoming:
Organizations can no longer rely solely on traditional cybersecurity strategies. AI-assisted defense systems, continuous monitoring, employee awareness training, and rapid patch management are becoming essential.
Cybersecurity experts also emphasize the importance of responsible AI development. Strong safety controls, ethical standards, and international cooperation may play a critical role in preventing advanced AI systems from becoming dangerous cyber weapons.
AI is reshaping cybersecurity at an extraordinary speed. What was once considered a future threat is now an active and growing reality. Reports from Google and other AI researchers suggest that cybercriminals are already leveraging powerful AI tools to automate attacks, identify vulnerabilities, and expand malicious operations worldwide.
At the same time, AI also offers opportunities for stronger cyber defense, faster threat detection, and improved security automation.
The challenge for governments, technology companies, and cybersecurity professionals will be ensuring that defensive innovation keeps pace with offensive AI capabilities.
As artificial intelligence continues to evolve, cybersecurity may become one of the defining battlegrounds of the AI era.
Source: The Guardian, May 11, 2026
