EPQ Presentation: Invisible Battlefields

Analysing AI-Powered Cyber Attacks and How to Stop Them

Slide 1: Title Slide

Invisible Battlefields: Analysing AI-Powered Cyber Attacks and How to Stop Them

Your Name EPQ 2025

Slide 2: Research Question

"How are AI-powered cyberattacks evolving, and how effective are current defensive strategies in mitigating their threats?"

Why this matters:

  • Average data breach costs: $4.88 million (IBM, 2024)
  • AI-driven attacks show 15% higher success rates
  • Attacks are faster, smarter, and harder to detect

Slide 3: What Are AI-Powered Attacks?

Traditional attacks: Scripted, predictable patterns

  • WannaCry (2017): Infected 200,000+ computers

AI-powered attacks: Adaptive and automated

  • Machine learning adapts to defenses in real-time
  • Generates thousands of unique attack variants
  • Learns from each attempt

Slide 4: Voice Cloning Attacks

How it works:

  • AI clones voices from 3 seconds of audio
  • Used in "vishing" (voice phishing) attacks
  • Impersonates CEOs, executives, trusted contacts

Real example:

  • 2019: Criminals used AI voice to steal €220,000
  • 50%+ success rate against untrained employees

Defense: Voice authentication, verification protocols

Slide 5: Deepfake Video Attacks

The threat:

  • AI creates fake video conference appearances
  • Real-time impersonation of CFOs, partners
  • Bypasses visual verification instincts

Real example:

  • 2024: Multi-million dollar fraud via deepfake video call

Detection challenges:

  • 65-70% accuracy for detection tools
  • Compressed video masks synthetic indicators

Slide 6: Social Media Targeting

AI-powered social engineering:

  • Analyses LinkedIn/social media profiles
  • Creates personalized phishing over weeks/months
  • 45% higher engagement than traditional phishing

How it works:

  1. AI scrapes your interests, career frustrations
  2. Crafts tailored job offers or opportunities
  3. Builds trust through consistent persona
  4. Introduces malicious links/attachments

Slide 7: Current Defenses - Organizations

Zero-Trust Architecture:

  • No implicit trust - verify everything
  • Limits lateral movement after breach

AI-Powered Tools:

  • Microsoft Defender for Cloud
  • IBM Security Operations Centers
  • Cisco Talos (processes 1.5 trillion daily events)

Success: $1.88M lower breach costs with AI deployment

Slide 8: Defense Limitations

The Problem:

  • AI attacks evolve faster than defenses
  • False positives cause "alert fatigue"
  • SMEs lack resources for advanced defenses
  • SolarWinds (2020): Even sophisticated defenses failed

The Reality:

  • Attackers have no ethical constraints
  • Defenders must balance security vs. usability
  • Supply chain = weakest link vulnerability

Slide 9: Key Findings

What I discovered:

✅ AI transforms attacks from scripted to adaptive ✅ Social engineering is the most dangerous vector ✅ Current defenses work BUT only for well-resourced organizations ✅ Human judgment still essential - AI can't replace it ✅ It's an ongoing arms race with no permanent winner

The gap: Small businesses create supply chain vulnerabilities

Slide 10: Conclusion & Future

The Answer: AI-powered attacks ARE evolving rapidly. Current defenses ARE partially effective, but gaps remain.

What's needed:

  • Hybrid AI-human defense strategies
  • Continuous adaptation, not one-time solutions
  • Global cooperation and threat intelligence sharing
  • Realistic expectations about AI limitations

Final thought: "AI-powered cyberattacks have redefined the invisible battlefield - only adaptive, ethical, and collaborative defenses can secure our digital future."

Presentation Tips:

  • Timing: ~2 minutes per slide = 20-minute presentation
  • Keep slides minimal - use bullet points, not paragraphs
  • Use visuals if possible (diagrams of attacks, screenshots)
  • Tell stories - the €220k voice fraud, deepfake video calls
  • Engage audience - ask if anyone's received suspicious LinkedIn messages
  • Practice - know your stats without reading slides

Comments

Post a Comment

Popular posts from this blog

Post-Implementation IT Satisfaction Survey – NexaTech Staff Dear Team, As part of our commitment to improving our IT infrastructure and security, we value your feedback. Please take a few minutes to complete this short survey. Your input will help us identify what is working well and where further improvements may be needed. 1. Onboarding & Training How confident do you feel using the new systems (Microsoft 365, OneDrive, Intune)? Very confident Somewhat confident Neutral Somewhat unconfident Not confident at all 2. Ease of Access How easy is it to access your files and applications remotely? Very easy Easy Neutral Difficult Very difficult 3. Device Performance Has the performance and reliability of your work device improved since the upgrades? Strongly agree Agree Neutral Disagree Strongly disagree 4. Security Awareness Do you feel more aware o...
Read more

task brief

Image
  Scenario You are a junior cyber security technician who has been hired by Shooting Star Drones Ltd to address some of the issues that it has recently experienced with its network and suspected unauthorised external access to its systems. Shooting Star Drones Ltd has been operating for 3 years and has a reputation for innovative and spectacular drone displays. The company aims to provide cheaper, more environmentally friendly drone displays to replace traditional fireworks. Displays can be of different sizes with displays often using up to 1000 drones. The company has 220 employees in its single office and is expecting to grow rapidly in the next 5 years.     Brief As part of your role, you are involved in a large security management project to identify and remediate the possible security breach. You have also been asked to support members of staff who requir...
Read more