Let's plug back in, Tech Insiders.

Search engines have started talking back, Meta is waving nine-digit checks at OpenAI devs, and the next Xbox wants a seat at your PC desk. Plus, 16 billion passwords are loose in the wild, and WormGPT has respawned with fresh villain skins.

Welcome to Monday. No coffee is strong enough for this one.

Here’s what you need to know today:

Before You Dive In:

The Neuron just dropped its revamped podcast—and it’s asking the big question: Will AI wipe out half of all white-collar jobs? We unpack the data, dodge the hype, and get smart insights from Microsoft’s Alexia Cambon on how to stay ahead in an AI-saturated workday.

Tune in on Spotify or wherever you get your podcasts! Or watch it on YouTube!

Gemini Gets a Mic: Google’s Voice-First AI Search Rolls Out

Google has introduced Search Live inside its AI Mode experiment, allowing US Labs users to hold a back-and-forth voice conversation with a custom Gemini model directly within the Google app on iOS and Android.

Tap the new Live icon, ask a question aloud, and the AI answers while surfacing links you can open later.

The session continues to run in the background, allowing you to switch to another app without losing context. A transcript button displays the exchange in text and allows you to type follow-ups. Google says camera-sharing—pointing your lens at something and asking about it—will be available "in the coming months," building on Live Lens demos from I/O 2025.

Competitors have also rushed to add voices: OpenAI launched its Advanced Voice Mode last year, Anthropic gave Claude vocal cords in May, and Apple continues to tease an LLM-powered Siri. Google's play, however, bakes the experience right into Search, the company's most valuable real estate.

Why it matters: If users stop tapping links and start talking instead, expect traffic to your favorite sites to plummet and the next frontier to be voice ads wedged between answers. It seems clicks are about to be an endangered species.

Share this newsletter!

Pulse Check

Would you actually talk to Google Search out loud?

Results from Wednesday's Pulse Check

Would ads in WhatsApp make you switch messaging apps?

Insider Intel

OpenAI–Meta Grudge Match: $100 Million Bonus Bait and Scale AI Breakup

What’s the price of a top AI brain? Meta’s opening bid is nine figures.

On the Uncapped podcast, OpenAI CEO Sam Altman revealed that Meta has dangled $100 million signing bonuses—plus similar yearly pay—to tempt elite OpenAI researchers into its new "superintelligence" unit.

Altman says none of his "best people" defected, but the situation highlights a rivalry that now stretches far beyond LinkedIn DMs.

The feud ratcheted up another notch after Meta bought a 49% stake in data-labeler Scale AI for $14.3 billion. Hours later, OpenAI began winding down its Scale projects, wary of piping proprietary data through a Meta-backed vendor. Google and xAI hit pause, too, leaving Scale contractors scrambling for work.

Bottom line: Expect the bonus bids to climb—and watch for potential lawsuits—because this OpenAI-Meta grudge is turning every talent negotiation into a nine-figure cage match.

Flash Bytes

Security Alerts

'Credentialpocalypse': 16 Billion Passwords Leak From Infostealer Dumps

Cybernews researchers uncovered 30 mega-datasets hosting a jaw-dropping 16 billion login credentials—fresh, structured, and weapon-ready for phishing and account takeovers.

The troves, briefly exposed via misconfigured storage buckets, cover everything from Apple IDs to corporate VPNs.

Because the dumps overlap, the actual victim count is unknown. But even a 0.5% success rate gives crooks millions of viable accounts. Researchers warn that the data will fuel ransomware, business email compromise attacks, and credential-stuffing barrages.

Pro tip: rotate your passwords, enable passkeys or 2FA everywhere, and maybe schedule that weekend Delete Old Logins party.

WormGPT Is Back… Now With Grok and Mixtral Under the Hood

Two new WormGPT variants—keanu-WormGPT (powered by xAI's Grok) and xzin0vich-WormGPT (running on Mistral's Mixtral)—have popped up on BreachForums, offering jailbreak-ready chatbots that happily churn out phishing kits and credential-stealing scripts.

Researchers at Cato Networks say worm-brand LLMs are becoming plug-and-play malware factories: simply drop in a commercial API, add a jailbreak prompt that suppresses guardrails, and rent access via Telegram. The moral: even closed-source models can mutate once they hit the underground.

A healthy reminder to review your email filters—and maybe teach Grandma what a 'jailbreak prompt' is (good luck).

Industry Shakeups

Microsoft Confirms the Next Xbox Will Be a Windows-Powered PC

In a short video, Xbox president Sarah Bond announced a fresh AMD silicon partnership and emphasized that next-gen Xbox consoles will run Windows and won't be locked to a single store. Think living-room PC that still plays your existing Xbox library, complete with AI-enhanced graphics.

Bond hinted at handheld form factors and backward compatibility, painting a future where "This is an Xbox" could apply to desktops, portables, and cloud clients alike. Translation: Microsoft wants you in the ecosystem, no matter the hardware.

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.

Advertise in Daily Tech Insider!

Daily Tech Insider is a TechnologyAdvice business

© 2025 TechnologyAdvice, LLC. All rights reserved.

TechnologyAdvice, 3343 Perimeter Hill Dr., Suite 100, Nashville, TN 37211, USA.

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