- The Rundown Tech
- Posts
- Binance hit by crypto’s biggest heist ever
Binance hit by crypto’s biggest heist ever
PLUS: No account, no Reddit
Good morning, tech enthusiasts. The biggest crypto heist in history just hit the biggest crypto exchange on Earth.
On Thursday, hackers drained $6.3B from Binance in minutes, and blockchain trackers say the trail leads to Pyongyang — raising the prospect that one of crypto’s most spectacular robberies could end up financing a nuclear-armed state.
In today’s tech rundown:
Binance loses $6.3B in record hack
Reddit locks out logged-out readers
Meta brings self-harm alerts to Facebook
Amazon rolls out five Alexa+ devices
LATEST DEVELOPMENTS
CYBERSECURITY

Image source: Vadim Art / Unsplash
The Rundown: Binance, the world’s largest crypto exchange, just lost about $6.3B on Thursday in the biggest crypto theft on record — and blockchain analysts are pointing at North Korea.
The details:
Binance suspects an engineer’s compromised computer enabled unauthorized withdrawals; it says there’s no evidence wallet keys leaked.
Crypto tracing firm TRM tracked $6.36B across 10 blockchains, mostly within 20 minutes. Roughly $3B has been frozen, though that doesn’t mean recovered.
Deposits are operating normally, and Binance expects withdrawals back within 48 hours. It promises to cover losses using emergency reserves and company funds.
TRM and Chainalysis point to North Korea’s TraderTraitor, which the FBI linked to February 2025’s $1.5 billion Bybit theft, roughly a quarter of this haul.
Why it matters: Binance’s pledge to reimburse customers could protect their balances while leaving stolen funds in the attackers’ hands. If North Korea is responsible, unrecovered money could support weapons programs that U.S. and allied governments say are partly financed through cryptocurrency theft.
TOGETHER WITH MULTIVERSE
The Rundown: AI models bill by the “token,” and those costs add up fast once AI is built into an app. CompactifAI API runs GLM 5.3, a powerful open AI model from Z.ai, for less than any provider tracked by independent benchmarker Artificial Analysis.
With CompactifAI API, you can:
Cut GLM 5.3 costs by 85% vs. Z.ai's own API
Go even lower with Quasar, Multiverse Computing’s compact version of GLM 5.3
Plug it into apps built for OpenAI, or into no-code n8n workflows

Image source: Brett Jordan / Unsplash
The Rundown: Reddit will eventually require an account to browse the site, CEO Steve Huffman said Thursday. The news came the same day the company reported $895M in Q3 revenue and admitted Google search traffic had slipped.
The details:
Huffman’s shareholder letter and a post on Reddit describe a “logged-in-only experience.” Tests start this quarter in select markets outside the U.S.
Search referral traffic was “down meaningfully” in Q3, and logged-out web visitors took most of the hit. Huffman blamed Google’s AI Overviews and AI Mode.
Huffman says signed-in users give Reddit more signals for targeting ads, engage more, and help the company control access to its posts and comments.
Third-quarter revenue jumped 53% to $895M, but Reddit’s fourth-quarter forecast of $1B to $1.03B fell short of Wall Street expectations.
Why it matters: For casual readers, this change means one more login screen between a Google result and the thread you want. As search sends fewer visitors, Reddit has less reason to leave its archive open to people who aren’t logged in, and more reason to push them toward creating an account for targeted ads.
META

Image source: Meta
The Rundown: Meta is bringing the self-harm search alerts it launched on Instagram in Feb. to Facebook and Messenger, and adding the same tripwire to teens’ conversations with Meta AI.
The details:
Parents who supervise a teen account get notified when the teen “repeatedly” searches for suicide or self-harm content “within a short period of time.”
Meta AI chats can also trigger alerts. Meta says every flagged chat is human-reviewed, with ambiguous cases defaulting to parent notification.
Alerts can arrive in-app, by email, text, or WhatsApp. Parents don’t see the teen’s exact searches or AI chats; they get guidance and support resources instead.
When Instagram launched the feature, the UK’s Molly Rose Foundation called it “flimsy,” warning it could alarm parents without fixing underlying safety issues.
Why it matters: The Facebook rollout comes amid mounting legal pressure over teen safety. Seven months ago, an LA jury found Meta and YouTube liable for addictive platform design; a New Mexico case has since left Meta facing $942M in penalties and five years of court-supervised safety reforms. Meta is appealing.
AMAZON

Image source: Amazon
The Rundown: Amazon is stuffing Alexa+ into a fresh wave of screens, speakers, clocks, and even a photo frame, betting that generative AI can finally make the smart home feel less like a graveyard of half-used gadgets.
The details:
The Echo Show 16 fits a 16-in. 1080p screen into a frame ~40% thinner than the Show 15’s, plus Fire TV, a 13MP night-vision camera, and Visual ID up to 17 ft.
The $249.99 Echo Studio adds a camera and LED for cloud streaming. The $99.99 Echo Dot doubles as a smart home hub and eero extender.
Two new devices: Echo Time, a $199.99 bedside clock, and the Alexa Photo Frame, a 10-in. touchscreen that pulls from Amazon Photos.
The same day Amazon unveiled its new Alexa hardware, it cut about 250 Devices and Services jobs, citing efficiency and product-roadmap priorities.
Why it matters: Echo Show is Amazon’s fastest-growing Echo line, up more than 25% year over year, while the company says people are talking to Alexa+ two to three times more than before. The new hardware lands as Amazon wants to prove those gains can turn Alexa from an expensive smart-home experiment into a meaningful AI platform.
QUICK HITS
A hacker reportedly leaked records on 41K DHS employees and contractors, including ICE and CBP personnel, exposing contact details and links to surveillance operations.
DoorDash agreed to buy Instacart for $17B to grow its grocery business and become the default app for local shopping and delivery.
Amazon’s AI spending spree drove 12-month free cash flow to negative $42.4B, while earnings and revenue missed Wall Street’s forecasts.
Netflix hired former Apple Podcasts exec Steve Wilson as it expands its video podcast lineup and competes with YouTube for creators.
Instagram is reportedly offering select YouTube creators $50K–$500K to post long-form videos on its new “Features” format, aiming to take on YouTube.
Finnish police opened a criminal probe into Google’s Muhos data center after its subsidiary cleared 330 hectares of forest before completing the environmental review.
Anduril and Qualcomm unveiled BlueChip, a laptop for the military and government with ~90% U.S.- or allied-sourced parts and a Snapdragon chip made in Arizona.
Artemis I found radiation near the Moon only slightly higher than in low Earth orbit, and detectors showed they could track astronaut exposure on future missions.
Elon Musk posted on X that Neuralink’s Blindsight implant is ready from an engineering standpoint and is now waiting only on regulators.
General Motors will pay $220M to settle claims that it secretly collected and sold drivers’ detailed driving data to insurers without proper consent.
Read our last AI newsletter: Inside the spy plot around Nvidia’s CEO
Read our last Tech newsletter: Apple’s big week kicks off with iPad mini
Read our last Robotics newsletter: The self-driving future grabs a giant shovel
Today’s AI tool guide: Make ElevenLabs voiceovers sound more natural
That's it for today's tech rundown!We'd love to hear your feedback on today's newsletter so we can continue to improve The Rundown experience for you. |
See you soon,
Rowan, Zach, Shubham, Jennifer, and Nate — The Rundown’s editorial team


Reply