Cybersecurity Trends for 2025 and Beyond

Introduction

• Overview of cybersecurity predictions for 2025 and beyond, comparing with previous predictions from 2024.

Review of Predictions from 2024

Passkeys Over Passwords

• Increased adoption of passkeys: 4.2 million passkeys saved in a password management software.

• 1 in 3 users are now using passkeys, with more companies accepting them.

• Continuing trend expected in the future.

AI Phishing

• Rise of AI-generated phishing emails that are highly personalized and appear legitimate.

• Standard mistakes in phishing emails are decreasing as AI improves its capabilities.

Deepfakes

• Instances of deepfake technology leading to financial fraud:

  • Deepfake impersonation of a CFO resulted in a $25 million loss.

  • Deepfake robocalls during elections causing misinformation.

Hallucinations in Generative AI

• Examples of AI generating incorrect information: A chatbot miscalculating running pace but correcting itself upon prompting.

• Ongoing issues with inaccurate outputs but improving over time.

Securing AI Deployments

• Significant concern for businesses focusing on securing AI technologies.

• Clients often inquire about the protection of AI systems from attacks.

AI in Cybersecurity

• Potential applications of AI in cybersecurity, including chatbots for Q&A and incident tracking.

• AI can assist cybersecurity analysts with summaries and responses.

Predictions for 2025 and Beyond

Shadow AI

• Increasing informal use of AI tools by individuals without organizational approval.

• Risks include data leaks and misinformation.

Deepfakes

• Continued enhancement of deepfake technology, posing risks to businesses and governments.

• Legal challenges surrounding authenticity of video evidence.

Malware Generation

• Generative AI’s ability to write malicious code leading to increased attacks.

• A major retailer noted a seven-fold increase in attacks attributed partly to generative AI exploitation.

Attack Surface Expansion

• New AI components broaden the attack surface for cyber threats.

• Risks related to the potential poisoning of AI systems.

Prompt Injection Attacks

• Generative AI susceptible to prompt engineering exploits,

• OWASP identifies this as a primary threat to large language models.

Positive Uses of AI in Cybersecurity

• AI could provide expert recommendations without automating decisions.

• Importance of human oversight in AI-influenced cybersecurity responses.

Quantum Computing and Cryptography

• Quantum computers may threaten current cryptographic methods in the future.

• Urgent need for organizations to develop quantum-safe cryptography to prevent future vulnerabilities.

Conclusion

• Invitation for viewers to share their own predictions, emphasizing community insights on future trends in cybersecurity.

Cybersecurity Predictions: 2025 and Beyond

Introduction

This document provides an overview of cybersecurity predictions for 2025 and beyond, comparing them with previous predictions from 2024.

Review of Predictions from 2024

Passkeys Over Passwords

The continued adoption of passkeys has been notable, with 4.2 million passkeys saved in password management applications, and 1 in 3 users now utilizing them. This trend is expected to persist in the future as more companies adopt these security measures.

AI Phishing

There has been a rise in AI-generated phishing emails that are highly personalized and appear legitimate. Common mistakes previously found in phishing emails are decreasing as AI technology improves.

Deepfakes

The use of deepfake technology has led to significant financial fraud, including a notable incident where a deepfake impersonation of a CFO resulted in a $25 million loss. Additionally, deepfake robocalls during elections have contributed to misinformation.

Hallucinations in Generative AI

Generative AI continues to generate incorrect information; for instance, a chatbot miscalculated running pace initially but managed to correct itself upon being prompted. While ongoing issues with inaccurate outputs persist, improvements are being made over time.

Securing AI Deployments

Business concerns regarding the security of AI technologies are increasing, with many clients inquiring about how to protect AI systems from potential attacks.

AI in Cybersecurity

AI holds promise for various applications in cybersecurity, including the use of chatbots for Q&A and incident tracking, as well as assisting cybersecurity analysts with summaries and response management.

Predictions for 2025 and Beyond

Shadow AI

The informal use of AI tools by individuals, without organizational approval, is on the rise, leading to risks such as data leaks and misinformation.

Deepfakes

Deepfake technology is expected to continue advancing, which poses risks to businesses and governments, exacerbated by ongoing legal challenges surrounding the authenticity of video evidence.

Malware Generation

Generative AI's capability to write malicious code is resulting in an increase in attacks, with a major retailer reporting a seven-fold rise in attacks, partly attributed to generative AI exploitation.

Attack Surface Expansion

New AI components are broadening the attack surface for cyber threats, raising concerns about potential poisoning of AI systems.

Prompt Injection Attacks

Generative AI is vulnerable to prompt engineering exploits, which OWASP has identified as a primary threat to large language models.

Positive Uses of AI in Cybersecurity

AI could offer expert recommendations in cybersecurity while emphasizing the importance of human oversight to prevent fully automating critical security decisions.

Quantum Computing and Cryptography

In the future, quantum computers may pose a threat to existing cryptographic methods, highlighting the urgent need for organizations to develop quantum-safe cryptography to prevent vulnerabilities.

Conclusion

The document concludes with an invitation for viewers to share their predictions, emphasizing the importance of community insights on future trends in cybersecurity.