Cybercriminals don’t need to break into your network — they just need to impersonate you. In this episode, Jeremy Nelson, CISO at Insight, walks through how threat actors use public-facing content, social media, and even dating apps to build convincing personas and launch ransomware attacks. He shares a real-world $25M deepfake scam, explains how AI is both the weapon and the shield, and outlines the internal verification protocols every company should adopt.

🎧 Listen now and explore more: https://www.insight.com/en_US/what-we-do/expertise/cybersecurity.html

Jump right to…
00:00: Welcome/intro
06:17: How attackers start with social media
08:00: Deepfakes and impersonation risks
09:48: Why AI vs. AI is the new battleground
12:05: Internal verification is your best defense
14:37: The unexpected risk of dating apps
16:23: Password reuse and dark web exposure
17:31: What attackers do after recon
20:28: CEO email compromise story
21:50: Why lurking threats matter
23:05: Humans are the weakest link
24:18: Insider threats and employee bribery

When AI makes a mistake, who’s accountable — the developer, the user, or the system itself?

In this episode, Meagan Gentry, National AI Practice Senior Manager & Distinguished Technologist at Insight, unpacks the concept of agentic AI and how organizations can embed accountability into autonomous workflows.

From the “moral crumple zone” to use case feasibility mapping, Meagan shares frameworks for building trust and driving ROI with AI agents.

Jump right to…
00:00: Welcome/intro
03:12: What is agentic AI?
06:45: Why accountability matters now
09:30: Explainability vs. performance tradeoffs
13:10: Ownership and moral crumple zones
17:15: Mapping accountability across AI lifecycle
20:21: Empowering users with AI awareness
25:32: Human in the loop vs. human in command
27:24: What CEOs must ask before greenlighting AI
29:30: Who belongs at the AI strategy table
30:58: Culture shifts and trust in AI agents

🎯 Related resources:

• https://www.insight.com/en_US/content-and-resources/blog/the-truth-about-ai-agent-risks-and-what-to-do-about-them.html

• https://www.insight.com/en_US/content-and-resources/blog/6-high-impact-agentic-ai-use-cases-executives-should-champion-today.html

Learn more: https://www.insight.com/en_US/what-we-do/expertise/data-and-ai/generative-ai.html

Subscribe for more episodes.

Everyone — including the markets — have been buzzing and reacting to a headline about a new MIT Research Report. You’ve likely read or heard iterations of the headline: 95% of Gen AI pilots are failing. But there’s a huge asterisk to this statement.

In this episode of Insight On, CTO of Product Innovation Amol Ajgaonkar explains why leaders shouldn’t take the 95% stat too seriously — and what to learn from it.

Amol breaks down the “stratlock” trap (when organizations get stuck in endless strategy and never move to execution). He also shares why pragmatic leaders focus on measurable outcomes, how to avoid common pitfalls, and what it really takes to drive AI adoption that works for people and the business.

Key topics:

• The difference between strategy and execution in AI projects
• How to build AI strategies that deliver real, measurable results
• The importance of starting small, iterating, and learning fast
• What leaders can do to move from planning to progress If you’re responsible for technology decisions, budgets, or outcomes, this conversation is for you.

Listen for practical advice and real-world examples that will help you lead with clarity.

🔗 For more resources, case studies, and solutions, visit https://www.insight.com.

If you found this episode helpful, please subscribe, leave a review, and share it with a colleague. It’s how we grow the conversation and help more leaders make better tech decisions.

Chapters

00:00 – Welcome

06:09 - The reaction to the MIT report on AI pilot failures

10:52 - Addressing executive hype and strategy lock

15:59 - The importance of an iterative AI strategy

35:50 - Red light, green light: The AI asterisk edition

43:03 - Final thoughts on a pragmatic AI approach