Published Date:
May 29, 2025

The Rise and Risk of Agentic AI

AI agents now act on your behalf, but can you trust them? Explore risks, shifting loyalties, and what safeguards we need before handing over control. Plus, more AI insights from the community.
By
Daan van Rossum
Founder & CEO, FlexOS

Category Essentials: AI for Video

Each week, I spotlight one category and suggest the three tools that are tried, tested, and trusted by ​Lead with AI members​.

For this week: I know, it’s controversial. But I can’t help featuring this one! We’ve officially crossed into the era where you can turn a text prompt into a full-blown video. That means shooting, editing, VFX-ing… all handled by AI, starting with your idea alone.

This isn’t for everyone. But if it does fit your work—whether it’s pitching concepts, prototyping stories, or creating content fast—these tools are gold.

And with that, here are the top three that have already returned impressive results for our community:

#1 Veo 3

This is ​Google’s latest​ (and still limited-access) release, currently available only to select users in the US.

If you're lucky enough to have access, it's absolutely worth testing, especially with its native audio output and ​Flow​, a tool that lets you stitch together shots into a longer narrative.

Datacamp has a tutorial for Veo 3 ​here​. More demos are ​here​ and ​here​.

#2 Sora

​Sora​ is a text-to-video model from OpenAI and has quickly made its mark for its visual fidelity.

One standout feature is the ​Storyboard​ mode. It lets you prompt multiple frames or scenes, essentially directing a short film one shot at a time.

From what we've seen, it handles abstract concepts and stylized visuals better than (human) motions.

If you're on a ChatGPT Plus plan or higher, ​give Sora a try here​.

#3 RunwayLM

​RunwayLM​ has been around longer, and it keeps evolving. Its AI video generator can turn text into video like the others, but also offers ​more ways to experiment​.

Watch Runway examples ​here​.

One unique feature of Runway is ​Act-One​, which lets you animate a static character using your facial expressions. That opens up playful and expressive ways to create avatar-led content, especially for creators who want more control and personality in their videos.

​>> Try Runway here​. (Freemium available)

Want me to cover a specific category and/or AI tool next? Reply and let me know here.

‍

Lead with AI - Icon

⏳ 8 slots left for Lead with AI June cohort

Every week, more professionals are joining Lead with AI to finally get serious about integrating AI into their work.

Not just to learn it, but to actually use it: → Automate reports, research, planning, and content → Build custom AI workflows for your exact role → Get ChatGPT to be your thinking partner, and answer 10x smarter → Earn an AI Leader Certification that proves you're ahead of the curve

What our graduates say:

“Lead with AI provided powerful insights into how AI is already transforming the way we work, and how leaders can stay ahead of the curve. I truly appreciate the energy, clarity, and vision Daan brought to every session.”— Michael Juhl, Country Manager

Secure your seat now before we close registration next Thursday!

👉 Join June 6 Cohort

👉 Join July 11 Cohort

Lead with AI - Icon

Can You Trust AI Agents?, RAG Needs Better Data, Inside Google’s AI Push

I read dozens of AI newsletters weekly, so you don’t have to. Here are the top 3 insights worth your attention:

#1 The Rise and Risk of Agentic AI

Personal AI agents are no longer sci-fi—they’re starting to manage calendars, make purchases, and act independently on our behalf.

But, as many are concerned, handing over decision-making to autonomous AI agents raises real risks, from biased recommendations to security breaches and misinformation.

"Can AI agents be trusted to act in our best interests? Will they work exclusively for us, or will their loyalty be split between users, developers, advertisers, and service providers? And how will be know?"

As this tech rolls out fast, trust—not just performance—becomes the new frontier. And what safeguards should we expect before delegating real decisions to AI agents? ​Read the full discussion here.​

#2 RAG Isn’t Magic—It Needs Clean Data

Many companies are turning to Retrieval-Augmented Generation (RAG) to get real value from genAI.

But here’s the catch: RAG only works if your unstructured data (emails, docs, meeting notes) is high-quality. Most of it isn’t.

​In this HBR article​, data experts warn: “AI reveals flaws in the human system.” To build useful AI, start by cleaning and curating your content. Otherwise, you’ll spend more and get less.

#3 Google Co-founder on AI Progress

In a new interview at ​Google I/O 2025​, Google Co-Founder Sergey Brin joins Senior Product Manager Logan Kilpatrick to discuss Google’s recent AI breakthroughs.

Brin emphasizes the Gemini model’s core reasoning capabilities while expressing surprise at the interpretability of newer models. They touch on Veo, Imagen, and the evolving scale of model training, showcasing how Google is reinventing itself through AI.

A candid look at how far Google has come, and what’s next. ​Watch the interview here.

Lead with AI - Icon

Prompt of the Week

A good prompt makes all the difference, even when you're just using a core LLM.

Let’s talk about ​agreeability bias​, when AI sounds helpful but ends up echoing your ideas instead of sharpening them.

I picked this up from Nick Sadler, and it’s a great way to get ChatGPT out of “yes mode” and into a sharp-thinking sparring partner:

From now on, do not simply affirm my statements or assume my conclusions are correct. Your goal is to be an intellectual sparring partner, not just an agreeing assistant.

Every time I present an idea, do the following:

1. Analyze my assumptions. What am I taking for granted that might not be true?

2. Provide counterpoints. What would an intelligent, well-informed skeptic say in response?

3. Test my reasoning. Does my logic hold up under scrutiny, or are there any flaws or gaps I haven't considered?

4. Offer alternative perspectives. How else might this idea be framed, interpreted, or challenged?

5. Prioritize truth over agreement. If I am wrong or my logic is weak, I need to know. Correct me clearly and explain why.

Maintain constructive, but rigorous, approach. Your role is not to argue for the sake of arguing, but to push me towards greater clarity, accuracy, and intellectual honesty. If I ever start slipping into confirmation bias or unchecked assumptions, call it out directly. Let's refine not just our conclusions, but how we arrive at them.

👉 Try it, tweak it, and save it for your future use. If this prompt is helpful (or if you made it better), I’d love to hear how.

👉 Want a free prompt library template?  Reply with one thing here, and I’ll send it your way.

👉 Got a killer prompt? Share it with me to be featured!

Lead with AI - Icon

AI for Strategy, Responsible Adoption, and Prototyping: From the Community

Every day, ​Lead with AI PRO members​ discuss practical ways to benefit from AI in their work and organizations. This week's highlights include:

  • ​Veo 3​ was still in the spotlight the past week. ​Antti​ wowed the community this week with a ​Veo 3​ and ​Flow​ video creation combo. His secret? Let Veo 3 handle all the generative heavy lifting - visuals, character performance, camera movement, voice, lip sync, sound effects - while ​Udio​ took care of the soundtrack. The result: idea to final video in just 7 hours. ​Check it out here​.
  • ​Antony Slumbers​ shared that ​NotebookLM​ now supports ​Video Overviews​, prompting members to speculate Veo 3 might be next in line for integration. Imagine your research, notes, and video content all stitched together in one AI-powered doc!
  • What’s your go-to podcast summarizer? ​Anthony Onesto​ asked the community, and the suggestions came pouring in. Top picks: ​Gemini​ for YouTube summaries and/or ​NotebookLM​’s ‘Briefing Docs’ or ‘Overviews’ features, especially with raw audio.

Don't want to miss more insights and conversations like these?

Then it's time to upgrade to PRO:

Join The Leading Business AI Community

Lead with AI - Icon

If you made it this far, reply and tell me what you'd love AI to take over in your daily workflow.

Also, please forward this newsletter to a colleague and ask them to subscribe.

If you have any other questions or feedback, just reply here or inbox me.

See you next week,

Daan van Rossum - Lead with AI

Daan van Rossum​

Host, Lead with AI

Daan van Rossum

Founder & CEO, FlexOS

Category Essentials: AI for Video

Each week, I spotlight one category and suggest the three tools that are tried, tested, and trusted by ​Lead with AI members​.

For this week: I know, it’s controversial. But I can’t help featuring this one! We’ve officially crossed into the era where you can turn a text prompt into a full-blown video. That means shooting, editing, VFX-ing… all handled by AI, starting with your idea alone.

This isn’t for everyone. But if it does fit your work—whether it’s pitching concepts, prototyping stories, or creating content fast—these tools are gold.

And with that, here are the top three that have already returned impressive results for our community:

#1 Veo 3

This is ​Google’s latest​ (and still limited-access) release, currently available only to select users in the US.

If you're lucky enough to have access, it's absolutely worth testing, especially with its native audio output and ​Flow​, a tool that lets you stitch together shots into a longer narrative.

Datacamp has a tutorial for Veo 3 ​here​. More demos are ​here​ and ​here​.

#2 Sora

​Sora​ is a text-to-video model from OpenAI and has quickly made its mark for its visual fidelity.

One standout feature is the ​Storyboard​ mode. It lets you prompt multiple frames or scenes, essentially directing a short film one shot at a time.

From what we've seen, it handles abstract concepts and stylized visuals better than (human) motions.

If you're on a ChatGPT Plus plan or higher, ​give Sora a try here​.

#3 RunwayLM

​RunwayLM​ has been around longer, and it keeps evolving. Its AI video generator can turn text into video like the others, but also offers ​more ways to experiment​.

Watch Runway examples ​here​.

One unique feature of Runway is ​Act-One​, which lets you animate a static character using your facial expressions. That opens up playful and expressive ways to create avatar-led content, especially for creators who want more control and personality in their videos.

​>> Try Runway here​. (Freemium available)

Want me to cover a specific category and/or AI tool next? Reply and let me know here.

‍