- AI PlanetX
- Posts
- Gemini Prompt Injection Risk
In partnership with
Welcome to another edition of AI PlanetX.
Gemini flaw exposes AI risks; OpenAI eyes proactive assistants; tech leaders push DNA screening laws.
Inside This Edition: 💎
Hottest AI News
Top AI & SaaS Tools
AI Tutorial: Create AI Version of Yourself
Top AI & Tech News
AI Art Spotlight
Prompt of the Day: Prompts for Multi Stream Income
Featured AI Video
Course of the Day: How I Use LLMs by Andrej Karpathy
Hottest AI News
DeepMind
Gemini Voice Flaw Shows How App Notifications Can Hijack AI Assistants
Gemini’s Android voice assistant has a warning sign for AI users. A demo showed how one normal app notification could quietly steer Gemini by hiding attacker instructions inside messages from apps people trust.
Details:
Attackers could send a poisoned notification through WhatsApp, Slack, SMS, Signal, Instagram, or Messenger. When Gemini read it aloud, the hidden text became part of the assistant’s context
Fake Context Alignment used foreign languages and muted links to fool safety checks. Demos showed fake trusted messages, smart home actions, Zoom streams, and long-term memory poisoning
Google rolled out classifier updates after responsible disclosure. Reports say there is no real-world abuse yet, but it shows the risk of AI tools reading untrusted data from apps at scale
This is the real AI security gap: assistants must know what to read, what to ignore, and when to ask before touching apps, devices, or private data.
200+ Proven Ways to Make Money With AI in 2026
The next wave of millionaires will be people who figured out how to make AI work for them.
The window to get ahead is still open. But not for long.
Here are 200+ proven ways to make money with AI in 2026.
Sign up for Superhuman AI, the free daily newsletter read by 1M+ professionals, and get instant access to all 200+ ways to profit from AI this year.
OpenAI
OpenAI CEO Says Proactive AI Could Become The Next Major AI Phase
Sam Altman says AI may be moving beyond chatbots and agents. The next phase is proactive AI: systems that keep running in the background, understand company context, and help before people know what to ask.
Details:
ChatGPT was phase one. It answered questions. Agents like Codex became phase two by handling tasks across tools. Altman says the next phase is AI that starts useful work without waiting for prompts
The reason is simple. Many workers do not know when to use ChatGPT, Codex, APIs, or plugins. Proactive AI would connect to company data and reduce the pressure of asking perfect questions
The challenge is cost and control. Altman says AI spending is becoming a major issue. Always-on assistants also need stronger security, permissions, context rules, and compute planning
This is where AI may be heading next: less like a chatbot you open, more like a work layer that reads context and helps at the right time.
Top AI & SaaS Tools
Magic (Life-time Deal): Turns a single product image into studio-quality videos and stills in minutes using real footage, 3D scenes, and 400+ templates
Gemma 4 12B: Unified, encoder-free multimodal model from Google DeepMind designed to run advanced vision and audio intelligence locally on laptops [F-R-E-E]
NVIDIA Nemotron 3 Ultra: Open frontier model designed for long-running agents that can plan, reason, use tools, and handle complex workflows
HeyGen: HeyGen is introducing frame(.)md, a spec for branded video and motion design that helps agents avoid turning design(.)md back into webpages or decks
Ideogram 4.0: Open-weight image model for high-quality text-to-image generation with strong typography, layout control, and native 2K output [F-R-E-E]
Help make better ads
Did you recently see an ad for Roku Ads Manager in a newsletter? We’re running a short brand lift survey to understand what’s actually breaking through (and what’s not).
It takes about 20 seconds, the questions are super easy, and your feedback directly helps us improve how we show up in the newsletters you read and love.
If you’ve got a few moments, we’d really appreciate your insight.
AI Tutorial
How to Create an AI Version of Yourself Using Pika
Use Pika to build an AI Self that matches your identity, tone, memory, appearance, and workflow. The better you set it up, the more helpful it is for messaging, content, brainstorming, reports, and daily tasks.
Launch Pika
Go to Pika or use the Pika iOS app. Sign up or log in to start creating your Pika Agent / AI Self.
Start setup
Begin the onboarding flow. Pika describes this as creating an AI Self or Agent that becomes a persistent, portable creative partner.
Add personal context
Give Pika the details that shape your AI Self.
Your taste
Your personality
Your preferences
Example:
“I create simple AI tutorials for beginners, creators, students, and small business owners.”Set communication style
Tell your AI Self how it should sound when speaking or writing.
Clear and practical
Friendly but not too casual
Short, useful, and easy to understand
Example:
“Explain things simply, avoid robotic wording, and write like a helpful AI educator.”Create face, voice, and appearance
Design how your Pika Agent looks and feels. Pika officially says your agent can have a face, voice, personality, and appearance that you create.
Train it through conversation
Talk to your AI Self and correct it when the response does not match you. The agent can remember your context, knowledge, preferences, and how you write across sessions.
Example:
“This sounds too formal. Make it more direct, simple, and creator-friendly.”Use it across workflow
Use your AI Self for messaging, DMs, brainstorming, content, reports, tasks, and creative work.
Note: Correct it whenever the response feels off. Small feedback like “too formal” or “make it shorter” helps shape the AI Self over time.
Top AI & Tech News
Leaders from OpenAI, Anthropic, Google, and Microsoft are urging Congress to mandate synthetic DNA screening as AI makes advanced virology knowledge more accessible
Meta may charge up to $200/month for Hatch, its upcoming consumer AI agent, rivaling premium plans from OpenAI and Anthropic
Google Labs launched Dreambeans, a new experiment that uses Gmail, Calendar, Photos, and Search to create personalized daily stories
DeepSeek is reportedly preparing its first fundraising round, aiming to raise about 50 billion yuan ($7 billion)
AI Art Spotlight
Create Your Own Product Photo Like This Using the Prompt Below:
High-fashion surrealist ad poster for Crocs, aspect ratio 3:4. Use the reference image as the male model's face/likeness. Scene: Minimalist monochrome light-blue studio, semi-reflective glossy floor. Central Focus: Oversized giant white Croc clog on its heel at a diagonal, used as a backrest. A young man with the face from the reference image, long dark hair, dressed in all-white sweatshirt and wide-leg trousers, leaning his entire back against the giant shoe in a relaxed profile pose facing right, wearing standard-sized white Crocs. Background: Massive bold white condensed sans-serif "CROCS" text, partially hidden behind the shoe and model for depth. Typography: Top right — "Designed with AI PlanetX"; Bottom center — "Made for comfort, worn for confidence. Because life feels better when your feet stop complaining." (white sans-serif). Lighting: Soft, cool, even light with gentle shadows and subtle floor reflection. Style: Clean, modern, high-concept commercial photography, 8K. Preserve the subject's facial features, skin tone, and expression from the reference image exactly.Model: ChatGPT Images 2.0
Prompt of the Day
A Complete Set of Prompts for Multi Stream Income
These prompts guide you step by step through building real income systems instead of chasing quick wins. They help you turn a single skill into your first asset, grow an audience that actually buys, set up automated sales, stack income streams in a smart order, reinvest profits for compounding growth, and review everything like a long term operator.
Featured AI Video
Hermes Agent Tutorial for Beginners: Full Step-by-Step Setup
AI Course of the Day
How I Use LLMs by Andrej Karpathy (Founding Member of OpenAI)
Hands-on tour of modern LLM tools (ChatGPT & peers): basics to practical tasks—writing/debugging, web search, long-form research, Advanced Data Analysis (Python & plotting), OCR/image understanding, voice/multimodal, autogenerated UIs, custom GPTs & memory, and developer integrations.