AI in Cybersecurity: Defender or Attacker’s Best Friend?

AI in Cybersecurity: Defender or Attacker’s Best Friend?

By Bert Amodol
Posted in Security
On March 25, 2026

AI is everywhere in 2026. It writes your emails, summarizes your meetings, generates your vacation photos (that you didn’t actually take), and occasionally makes you wonder if it’s quietly coming for your job. But in cybersecurity, AI isn’t just helpful, it’s chaotic neutral. Because while defenders are using AI to stop attacks, attackers are using the same technology to get better at launching them.

So, the real question is:

Is AI your security team’s new MVP… or the hacker’s ultimate sidekick?

The Good: AI as Your Overachieving Security Analyst

Every security team has that one person who never sleeps, catches everything, and somehow drinks five coffees before 9 a.m. AI is that person. Minus the HR complaints.

Threat Detection at “How Did You Even Catch That?” Speed

AI can scan ridiculous amounts of data in real time. Logs, traffic, user behavior, all of it. While a human analyst is still opening their third dashboard tab, AI is like:

“Hey, that login from another country 8 seconds ago? Yeah, not great.”

Instant Response (No Meetings Required)

AI doesn’t schedule a call to discuss next steps. It just acts.

  • Suspicious activity is blocked.
  • A compromised device is isolated.
  • Weird behavior is flagged.

No calendar invite. No “circling back.” Just action.

Predicting Attacks (Basically Cybersecurity Weather Forecasting)

AI can spot patterns and trends that hint at future attacks.

Is it perfect? No.

Is it better than guessing? Absolutely.

Think of it as, “There’s a 70% chance of ransomware with a slight chance of panic.”

The Bad: AI Just Joined the Hacker Group Chat

Here’s the problem: attackers also discovered AI, and they got very excited.

Phishing Emails Went From “LOL Nice Try” to “Wait… Is This Real?”

Remember when phishing emails were full of typos and weird grammar? Those days are gone. AI now writes phishing emails that are polished, personalized, and slightly too convincing.

That “quick favor” email from your boss? Yeah… might not be your boss.

Automation: Hackers Now Have Unlimited Interns

AI lets attackers automate everything.

  • They can scan for vulnerabilities.
  • They can test stolen passwords.
  • They can craft social engineering messages.

It’s like giving hackers an army of unpaid interns who never sleep and never complain.

Deepfakes: Because Reality Needed More Plot Twists

AI can now mimic voices, faces, and writing styles. Which means:

  • They can generate fake Zoom calls.
  • They can generate fake voicemails.
  • They can generate fake executives asking for very real wire transfers.

At this point, even your mom might need a safe word to prove it’s really her.

So… Who’s Winning?

Honestly? It depends on the day. AI is making defenders faster and smarter, but it’s also making attackers more efficient and scalable.

The real shift is this:

You don’t have to be a genius hacker anymore, you just need good tools, and AI is the ultimate tool.

What This Means for You (Yes, You)

Whether you’re running a company or just trying not to get your email hacked, here’s the deal:

Humans Are Still the Weakest Link

AI can write the perfect phishing email…

…but it still needs someone to click it. Don’t be that person.

Speed Is Everything

Attacks happen faster now. If your security is slow, AI-powered threats will absolutely lap you like it’s the Daytona 500.

Layered Security = Not Getting Fired

No single tool will save you. You need layers:

  • Enable MFA (seriously, turn it on).
  • Use endpoint protection.
  • Monitor your environment.
  • Train users and build security awareness.
  • Adopt AI-powered security tools.

Think of it like cybersecurity lasagna. More layers = better outcome.

Final Thoughts: Welcome to the AI Arms Race

AI isn’t good. AI isn’t evil. AI is just… really, really effective. And right now, both sides are using it. So the difference comes down to this:

Are you using AI better than the people trying to break into your stuff?

Because they’re definitely using it. If you’re not thinking about AI in your security strategy yet, you might already be behind.

No pressure.

Bert Amodol

Bert Amodol

Bert Amodol is cybersecurity fanatic with over 25 years of experience in information security and compliance. In his current role as Principal Architect at Gotham Technology Group he assists customers in understanding their cybersecurity risks and determining current gaps as well as options to bridge them whilst integrating new solutions with their existing security architecture. In his prior role as Director of Citi’s Endpoint and PKI Security Engineering he managed teams which provided the endpoint security and PKI standards for Citi’s global infrastructure. He is also actively involved in defining strategic security direction and security policies and standards.