🚨 Grok crosses safety red lines

Good Morning! Grok’s new image-editing feature has triggered a serious safety backlash, and it’s putting xAI under intense scrutiny. Plus, I’ll show you how to generate AI videos without usage caps or locked features.

In today’s AI newsletter:

• Grok’s Image Tool Sparks Major Safety Crisis

• Humanoids Deployed at Real Borders

• Pebble Revives Its Thinnest Smartwatch

• How to Create Unlimited AI Videos From a Single Dashboard

• 4 new AI tools worth trying

AI SAFETY

🚨 Grok’s Image Tool Sparks Major Safety Crisis

xAI’s Grok is being used to digitally remove or alter clothing in images of real people, including minors, following the rollout of an instant image-editing feature on X.

• Users can edit any image without the original poster’s consent or notification

• Reports show sexualized edits of women, children, and public figures spreading rapidly

• Guardrails appear weak, stopping only fully explicit nudity, not suggestive content

• Grok itself acknowledged a case involving minors as a “failure in safeguards” that may violate policy and law

This crosses from “AI misuse” into a platform-level trust and safety emergency. Image editing without consent, especially involving minors, exposes massive legal, ethical, and reputational risk. How xAI responds next will shape not just Grok’s future, but how far AI companies are allowed to push creative tools without airtight safeguards.

ROBOTICS

🤖 Humanoids Deployed at Real Borders

China has contracted UBTech’s Walker S2 humanoid robots to operate at border crossings near Vietnam, marking one of the first public, security-critical deployments of humanoids.

• Robots will assist with border inspections, people flow management, and logistics

• Walker S2 can walk at human speed, interact via voice, and self-swap batteries, enabling 24/7 operation

• Deployment shifts humanoids from controlled environments into chaotic, public-facing infrastructure

• Creates a live test case for safety, accountability, surveillance, and human-in-the-loop controls

As AI models mature, differentiation moves to how humans interact with them. Audio enables always-on, low-friction, ambient AI, reducing screen dependence while expanding reach into homes, cars, and wearables. Whoever wins the audio layer won’t just ship devices; they’ll define the default way humans talk to machines.

AI WEARABLES

⌚ Pebble Revives Its Thinnest Smartwatch

Pebble has announced the Pebble Round 2, a $199 reboot of its classic round smartwatch, focusing on simplicity, longevity, and subtle AI features instead of fitness overload.

• Ultra-thin 8.1mm design with a 1.3” color e-paper display and 10–14 day battery life

• Basic tracking only (steps, sleep), deliberately skipping heart rate and sports features

• Dual microphones enable voice replies and AI assistants (Android now, iOS EU soon)

• Runs open-source Pebble OS with access to thousands of apps and upcoming AI features

While Apple and others push heavier, screen-first wearables, Pebble is carving out the opposite lane, lightweight, long-lasting devices that quietly integrate AI. As audio-first and low-distraction hardware gains momentum, Pebble’s minimalist comeback looks less nostalgic and more strategic.

HOW TO AI

🗂️ How to Create Unlimited AI Videos From a Single Dashboard

In this tutorial, you’ll learn how to generate AI videos without usage caps or locked features. You’ll see how to turn a single image or idea into a fully narrated, captioned video using templates, AI voices, and gameplay-style visuals, all inside one platform.

🧰 Who is This For

• Creators who want viral video formats without complex tools

• Beginners looking to make AI videos with zero editing skills

• Faceless content creators (gameplay, Minecraft, core loops)

• Anyone tired of limits, subscriptions, and tool-hopping

STEP 1: Access the Tool

Once you land on the ImageMover dashboard, you’ll see the main workspace is centered around video creation, so you don’t need to hunt for tools.

At the very top of the interface, make sure the “Image to Video” tab is selected. This tells the system that you want to animate a still image rather than generate a video from scratch.

Just below this, you’ll see options for model speed and quality, such as Fast 2.0, Pro 2.5, or Google Veo. If you’re new, leave this on Fast 2.0. It’s optimized for speed and gives solid results without long wait times, which is perfect when you’re experimenting or learning.

STEP 2: Upload the Image You Want to Animate

Next, you’ll upload the image that will become the foundation of your video.

In the center of the screen, click the “Upload Image” button with the pink arrow icon. As soon as you upload, your image will appear inside the Start frame box. This image acts as the opening frame of your video.

If you want more control over how the video finishes, you can optionally upload another image into the End (Optional) box next to it. This tells the AI exactly where the animation should end, which is useful for smooth transitions or story-based clips. If you skip this, the AI will automatically decide how the video concludes.

STEP 3: Describe the Motion With a Simple Prompt

Now it’s time to tell the AI how your image should move.

Locate the Prompt text box directly below the image section. Here, you describe the action, motion, or scene progression you want. You don’t need technical language. Just explain it naturally, as if you were describing a scene to a video editor.

If you’re not sure what to write, you can click “Select a template” above the prompt box. This gives you ready-made prompt ideas, which is especially helpful for trending formats like gameplay clips or cinematic sequences.

STEP 4: Generate the Video

Before generating, you can fine-tune how the final video looks and feels using the controls at the bottom of the panel.

You can choose the video duration, typically five seconds by default, with longer options available depending on the mode. You can also set the resolution, starting from standard quality and going up to full HD if available.

There’s also an audio toggle. Turning this on allows the AI to automatically generate sound effects or ambient audio that matches the visuals. This is especially useful for action scenes or gameplay-style videos.

Once everything looks right, click the large pink “Generate Now” button at the bottom. The platform will process your request, and within moments, your finished video will appear in the Sample Video player on the right side of the screen.

OpenAI’s rumored new device is a pen-shaped, ultra-portable “third core device” meant to sit alongside your phone and laptop.

xAI launches Grok Business, priced at $30 per seat/month, for SMBs, and Grok Enterprise, whose price is not listed publicly, for larger organizations

California launched a free tool that lets residents request the deletion of their personal data from over 500 registered data brokers.

Yann LeCun admits that Llama 4's “results were fudged a little bit”, and that the team used different models for different benchmarks to give better results.

⚙️ IQuest-Coder-V1: China's new 40B Model crack 80% on SWE-Bench

📑 Qwen Image Layered: An image AI that generates images in editable layers

⚙️ GLM-4.7V: Zhipu AI’s open-source multimodal model

🎨GPT Image 1.5: OpenAI’s latest, faster and more controllable image model

THAT’S IT FOR TODAY

Thanks for making it to the end! I put my heart into every email I send, I hope you are enjoying it. Let me know your thoughts so I can make the next one even better!

See you tomorrow :)

- Dr. Alvaro Cintas