AI arms race: How AI will be used by cyber-attackers (and defenders)

It’s no surprise that AI’s explosive growth in the last five years has also greatly expanded the need for sophistication and preparation from security threats. While artificial intelligence presents new challenges, machine learning and neural networks also expand security teams’ footprints and ability to keep their companies and customers safe from cyberattacks. Harnessing the power of AI can help organizations detect and even predict threats faster than ever before.  


Read on for more about the state of AI in cybersecurity, threats and advantages IT professionals need to be aware of, and a deeper focus on password security in the AI arms race. 

How hackers are using AI against you

With the cost of data breaches rising year over year according to a new report from IBM, it’s more important than ever to understand how attackers are using AI to breach, phish, and attack inside organizations across the globe.  

Automated phishing and social engineering 

AI can generate highly convincing phishing and social engineering attacks by analyzing large datasets of human interactions. AI tools make it easier than ever for attackers to craft personalized and convincing messages that are more likely to trick victims into surrendering sensitive data. 

According to a recent article from Forbes, Business Email Compromise (BEC) attacks are presenting an increasingly large challenge to corporations and IT professionals thanks to convincing deepfakes that can persuade employees by posing as executives in email, text, and even Zoom.  

Enhanced malware 

Artificial intelligence can be used to create more sophisticated and adaptive malware. For example, AI can help malware evade detection by learning from security systems and adjusting its behavior in real-time. Evolving rapidly and with so many different signatures makes it harder for traditional security measures to identify and neutralize threats. 

Targeted attacks 

Between analyzing enormous amounts of data to target individuals or organizations and its vast automated capacity, AI’s ability to leverage targeted attacks at scale is unprecedented. These attacks used to require sophisticated networks and enormous computing and people power, but attackers can now tailor their attacks using AI models to maximize impact and success rates.  

Key applications of AI in cybersecurity 

The role of AI in cybersecurity is two-sided: While AI is intensifying security challenges at every turn, machine learning also provides huge advantages for IT professionals willing to embrace it.  

Advanced threat detection 

Hackers aren’t the only group that can mobilize AI’s ability to analyze large volumes of data in real-time to detect anomalies and potential threats that might be missed by human analysts. IT professionals can set up a robust network of machine learning algorithms to identify patterns and behaviors that indicate malicious activity to allow for faster and more accurate threat detection. 

Predictive behavioral analytics 

Machine learning dramatically increases the surface area of user and entity behavior analytics (UEBA) with comprehensive data analysis. By analyzing historical data and current trends, AI can parse typical behaviors while flagging anomalous patterns or even predicting future attacks to stop threats fast.  

Automated responses 

AI can automate responses to detected threats to reduce the time it takes to mitigate attacks. For example, AI systems can automatically isolate infected systems, patch vulnerabilities, and deploy countermeasures to minimize the damage caused by cyber-attacks. 

AI in cybersecurity: password deep-dive 

Now let’s focus on how attackers can use AI to their advantage when it comes to passwords as well as ideas for fighting back. 

Common AI threats to password security 

AI is putting up a strong threat to even the most secure password security programs. Here are a few of the most common ways hackers are using AI to their advantage. 

Password cracking 

Efficient and sophisticated password-cracking algorithms are more prevalent and impactful than ever. By analyzing large datasets of leaked passwords, AI can identify common patterns and generate more accurate guesses to make brute-force attacks extra effective. 

Social engineering 

Creating highly personalized and convincing phishing messages can tempt users into revealing their passwords. By analyzing social media and other online data to deploy more convincing personas, AI can craft emails, texts, phone calls or even live video chats that are more likely to trick users into surrendering sensitive information

The World Economic Forum ranked misinformation and disinformation as the top global technological threat in 2024 due to the enormous impacts they can have on geopolitical affairs. This misinformation can start with a simple email or text at work and ripple outwards. 

Credential stuffing 

AI can automate and optimize credential stuffing attacks, where stolen login credentials are used to gain unauthorized access to multiple accounts. Generative AI can quickly test large numbers of credentials across various platforms, increasing the likelihood of successful breaches. 

Continuous Scan Password Policy icon
Continuously block 4 billion+ compromised passwords in your Active Directory

How to use AI to strengthen passwords 

AI systems can be hugely beneficial for IT security teams looking to get ahead of cyber-attacks leveraging passwords. 

Behavioral biometrics and pattern recognition 

AI can be used to implement behavioral biometrics, which monitor how users type their passwords. By analyzing typing patterns, AI uses deep learning to detect anomalies that might indicate unauthorized access, even if the correct password is entered. For example, AI can detect unusual login patterns, mouse movements, or locations and require additional verification steps to make it harder for attackers to gain access with stolen passwords. 

Machine learning can also be used to detect those same patterns to make logging in simpler or even passwordless for employees when all appears secure.  

Password strength analysis 

Analyzing enormous data sets in the background is one of AI’s most universal use cases, and password strength is a perfect use of that powerful machine learning. By analyzing passwords against known patterns and vulnerabilities, AI can automatically identify weak passwords and prompt users to create stronger ones to enhance existing security protocols at scale. AI chatbots can also stand in for IT support in the case of weak or forgotten passwords to streamline getting users back on their feet securely.  

Multi-factor authentication (MFA) enhancements 

Using AI to analyze user behavior and login context can improve the effectiveness of multi-factor authentication. If a user’s location or keystrokes indicate a more risky login environment, AI can automatically increase MFA requirements from, say, receiving a text message to biometric confirmation.  

Secure your Active Directory against password attacks  

AI’s tangible threat to cybersecurity is on the rise and important for any IT security team to be paying attention to in 2025 and beyond. As attacks evolve, it’s key to make sure your Active Directory is secure and easy routes into your organization are closed off to attackers.   

With Specops Password Policy and Breached Password Protection, organizations can continuously protect themselves against over 4 billion more known unique compromised passwords. This includes compromised passwords that could be considered ‘strong’ and have been stolen by malware.   

Our research team’s attack monitoring data collection systems update the service daily and ensure networks are protected from real world password attacks happening right now. The Breached Password Protection continuously scans your Active Directory for breached passwords and allows you to alert end users with customizable messaging that helps reduce calls to the service desk.  

Interested in seeing how this might work for your organization? Have questions on how you could adapt this for your needs? Contact us or see how it works with a demo or free trial.

(Last updated on February 5, 2025)

picture of author marcus white

Written by

Marcus White

Marcus is a Specops cybersecurity specialist based in the UK. He’s been in the B2B technology sector for 8+ years and has worked closely with products in email security, data loss prevention, endpoint security, and identity and access management.

Back to Blog

Related Articles

  • Add users to an Active Directory group based on user attributes

    A while back I visited a company to help install Specops Password Reset. They wanted a Group Policy configured for password resets using SMS to be applied to users with a corporate mobile phone. All other users should be reached by a Group Policy configured for password resets using security questions. The best way to make…

    Read More
  • Creating a custom password-exclusion dictionary with ChatGPT

    When cybercriminals attempt to crack passwords, it makes sense to go for the lowest hanging fruit. They’re going to start by trying the most common, easy-to-guess passwords, as chances are some end users are bound to have chosen them. So it makes sense for organizations to use the same logic – block the weakest passwords…

    Read More
  • Passwords are the biggest threat to GDPR compliance

    Data protection authorities have imposed their first GDPR fine in Germany after a local chat service exposed 330,000 credentials. What will be the consequence of the Collections leak that impacts 2.4 billion usernames & passwords?

    Read More