PLUS: New 'Pixnapping' Flaw Lets Android Apps Steal Screen Data

Oct 15, 2025

 

Read in Browser 

 

Oct 15, 2025

Read in Browser

Power up, Tech Insiders.

Whether it's Salesforce wiring agents into every workflow or Figure teaching bots laundry etiquette, software is rewriting labor in real time.

Grab a front-row seat—your dishwasher may soon have a résumé.

Here's what you need to know today:

  • Figure 03 promises mass-market home robots
  • Salesforce rolls out Agentforce 360 enterprise AI
  • 'Pixnapping' attack leaks Android screen contents
  • Framework firmware flaw bypasses Secure Boot
  • California tightens AI child safety laws
Advertise in Daily Tech Insider

Figure 03: Household Robots for the Masses

Ford had the Model T; Figure wants the Model Tea-Towel.

Figure AI's new Figure 03 isn't just a shinier prototype; it's purpose-built for assembly lines targeting an eventual 100,000 humanoids over four years.

Powered by its Helix vision-language-action brain, the 5-foot-6-inch bot can load dishwashers, fold laundry, and recharge wirelessly through its feet. Upgrades include palm cameras, paper-clip-sensitive fingertip sensors, and a washable fabric "sweater." It's also 9% lighter than its predecessor.

Figure 03: Household Robots for the Masses

GIF via Figure

Backed by $1 billion from Jeff Bezos, OpenAI, Nvidia, and Microsoft, Figure is building a BotQ factory to ship an initial 12,000 units annually, with hopes for home readiness by 2026.

CEO Brett Adcock touts that the bot learned to fold towels from just 80 hours of video, but demos still drop socks, and rival roboticists question the humanoid form factor.

TIME named Figure 03 a 2025 Best Invention; skeptics call it a slick highlight reel, not unlike rival Tesla's Optimus, which can strike a kung fu pose but still can't reliably stock a shelf.

Even if Helix scales like ChatGPT, safety, pricing, and catch-the-falling-plate reflexes remain open tickets. For now, BMW's Spartanburg plant is the lone real-world test bed.

Why it matters: If Figure nails affordable general-purpose bots, labor economics, eldercare, and your weekend chore list get upended. If not, it's another viral towel fold—and you're still sorting socks.

Share this newsletter!

FacebookXLinkedIn

Pulse Check

Would you trust a humanoid helper like Figure 03 in your home?

😎 Absolutely, laundry day just got outsourced
🤔 Maybe, after the beta glitches are gone
😱 No way, I'll keep my socks fold-free

Results from Yesterday's Pulse Check

A company designing both the AI model and the chips it runs on makes you…

A company designing both the AI model and the chips it runs on makes you…

Insider Intel

Salesforce Debuts Agentforce 360 Platform

Salesforce just gave every bot its own Slack badge.

The CRM giant recently launched Agentforce 360, a unified stack that lets enterprises design, govern, and monitor AI agents across sales, service, marketing, and more.

Key pieces: Agentforce Builder (a no-code studio), Agent Script for if/then guardrails, a hybrid reasoning engine, and native voice.

Twelve thousand customers, from Reddit to Adecco, are already offloading routine tickets and calls. Slack becomes the "agentic OS," where humans and bots collaborate.

Salesforce Debuts Agentforce 360 Platform

Image via Salesforce

The bigger news? A landmark OpenAI deal means users will be able to query Salesforce records and build Tableau charts directly from within ChatGPT by year's end.

Partnerships with OpenAI and Anthropic also let customers plug their choice of next-gen GPT-5 and Claude models directly into the platform's Atlas reasoning engine.

Competition is fierce: Microsoft's Copilot Studio and Google's Vertex Agent Builder tout similar orchestration. Salesforce counters with deep governance, a new agent marketplace in Slack called AgentExchange, and a promise that AI elevates people, not replaces them.

Cue the joke about cloud bills: at least these agents don't demand free snacks... yet.

Flash Bytes

🤖 Hire a Vision Pro, Minus the Guesswork

The Computer Vision Engineer Hiring Kit packs a job spec, interview Qs, and advert template—saving you 20 hours of recruiting legwork.

 

📊 Dive Into Data for 87% Less

Grab The Advanced Data Science Bundle—eight expert courses (Power BI, TensorFlow, PySpark, and more) with 50.5 hours of lifetime training—for just $29.99 (87% off).

 

🔍 Audit Your Endpoints in One Dashboard

Snag the Best Remote Monitoring and Management Software 2025 list to automate patches and keep remote fleets hacker-free in 2025.

 

💻 This section contains sponsored tech insights. Advertise with us!

Security Alerts

New 'Pixnapping' Drains Android Secrets

A multi-university research team showed how a permission-free app can infer on-screen pixels via GPU side channels on top Google and Samsung phones, stealing 2FA codes in under 30 seconds.

Google issued a partial patch for CVE-2025-48561 in September, but a workaround still works; a full fix lands in December.

New 'Pixnapping' Drains Android Secrets

Image created with Gemini

For now, there are no known in-the-wild exploits. Until a full patch arrives, install updates, scrutinize sideloaded apps, and maybe keep Authenticator on a separate device.

Pro tip: Waving your phone frantically does not blur the pixels, only your dignity.

Framework Laptops Exposed to Shell Backdoor

Firmware firm Eclypsium found that roughly 200,000 Framework PCs ship with a signed UEFI shell containing the "mm" command—a tool that can directly edit system memory—enough to null Secure Boot, your PC's startup security check, and let an attacker with physical access stroll in with bootkits.

Models spanning nearly the entire lineup, from 11th-gen Intel to the latest Ryzen chips, are affected; BIOS updates removing the vulnerable command from the shell (and DBX revocations, basically a digital blocklist for bad code) are rolling out.

In the meantime, set a BIOS password or yank the Framework key.

Sometimes the most modular part is the security hole you didn't order.

 

Industry Shakeups

California Clamps Down on Kid-Unsafe AI

Governor Newsom just signed a package of first-in-the-nation laws aimed at shielding minors from algorithmic harm. SB 243 forces "companion" chatbots to age-verify, steer self-harm queries to hotlines, and remind kids every three hours they're chatting with code.

AB 56 slaps cigarette-style mental-health warnings on social media, while AB 1043 makes device OSes collect user ages and share brackets with apps—an approach even Google and Meta endorsed (though Apple sat this one out).

California Clamps Down on Kid-Unsafe AI

Image created with Gemini

Under the new rules, deepfake porn distributors now face up to $250,000 in civil penalties (AB 621), and tech firms can't blame rogue AI for damages (AB 316).

Supporters hail a workable template; Hollywood fears chaos for shared streaming profiles; industry lobbyists warn of innovation chill. Yet with DC gridlocked, California's guardrails may become the default for 40 million Americans and, by extension, everyone online.

So yes, starting around 2027, your phone will ask how old you are—again—and this time it's for the children.

Meet Our Author

Justin Meyers

Justin Meyers

Writer at TechnologyAdvice

Justin Meyers is an investigative writer and editor who draws on over a decade of meticulous hands-on research to deliver the full, trustworthy story behind consumer and enterprise tech, including cybersecurity.

 

Curious about where AI is really headed?

The Neuron cuts through the noise to bring you smart, hype-free takes on the latest AI trends, tools, and breakthroughs. Join 500,000+ professionals from top companies like Microsoft, Apple, Salesforce and more.

Subscribe now!
Sign Up
Contact Us

Advertise in Daily Tech Insider!

Daily Tech Insider is a TechnologyAdvice business

© 2025 TechnologyAdvice, LLC. All rights reserved.

TechnologyAdvice, 3343 Perimeter Hill Dr., Suite 215, Nashville, TN 37211, USA.

FacebookXLinkedIn

Want to change how you receive these emails? Update your preferences or unsubscribe here.