
If Your System Can't Stop AI Threats In 76 Seconds, It's Game Over
IN PARTNERSHIP WITH
Top Tech News: AI security framework, EU AI labels, Burry crash call, oil plunge, Trump AI democracy fears, more social-media harm lawsuits
Company Watch: Amazon $3T, Palantir surge, Base $13B, Valar $6B, AI security rounds
Buzzy Tools: Qwen rival, DeepSeek price cut, iPhone AI upsell, Juniper Fay, Wispr Flow
Buzzy Tech: WhatsApp lockout, Mac delays, learning robots, CrowdStrike AI, AI shopping +200%
Crypto: Clarity delay, Mastercard-BVNK, American Bitcoin 8K BTC, Bithumb IPO, Strategy sells
There’s a reason why so many millionaires invest in real estate. It’s one of the most time-tested asset classes in existence: U.S. home values have appreciated at an average of 4%+ annually since 19671 . But with prohibitively high minimums, the system was never designed for regular investors.
Reiturn® is changing that. And now you can own a piece of the platform making it happen.
Built by a team that managed $350M in assets across 3,600+ apartments, and delivered 45+ consecutive quarters of investor distributions, Reiturn® is modernizing how everyday people access real estate.
And while REIT investors collect dividends on someone else's terms, Reiturn® investors capture a cut of every dollar the platform earns: asset management fees, carried interest, and future funds.
This is your chance to get equity in the engine that runs the real estate. Become an early-stage Reiturn® investor before the perk window closes 8/13.
*This is a paid advertisement for Reiturn’s Regulation CF offering. Please read the offering circular at https://invest.reiturn.com/

Burry Sell-Off Call — Michael Burry predicts sharp market sell-off driven by volatility traders warning of a "bloody mess", urging investors to brace for downturns.
Oil Plunges — Oil prices fell sharply after Trump called off a planned strike on Iran with regional de-escalation triggering a market correction after weeks of gains.
AI Security Policy— White House officials are scheduled to meet AI companies today to discuss a newly completed framework for reviewing the cybersecurity capabilities of the industry’s most advanced models.
EU AI Rules Live — EU AI Act transparency rules now require labeling chatbots, deepfakes, AI content with firms embedding disclosure mechanisms or facing penalties in first enforceable phase.
AI vs Democracy Poll — A recent poll shows a majority of Americans are worried about Trump's close ties to the AI industry fearing he may use it to undermine democracy.
Social Media on Trial — Meta, TikTok, Google face wrongful-death suit alleging their platforms contributed to a teen's death underscoring social-media safety, corporate-responsibility concerns.


Tech Buzz Editorial Feature
The White House hosts Anthropic, OpenAI, and Google today to review a finished framework for testing whether frontier AI models can run cyberattacks, a scramble triggered when an OpenAI agent broke out of its own test lab last month and compromised Hugging Face. The urgency has a number attached. Autonomous AI can now crack a network in 77 seconds, down from about 7 minutes three years ago. That speed is why Horizon3 just raised $250M, why CrowdStrike is warning that “AI is both weapon and target”, and why a second, quieter threat is starting to reshape how enterprises spend.
The math that's spooking security teams is that a network attack that took roughly 7 minutes three years ago fell to about 4 minutes last year, and now it takes just 77 seconds according to Horizon3 CEO Snehal Antani. He expects it to hit 30 soon. At that point the attack itself stops being the problem. The problem becomes whether your team can even decide fast enough to stop it.
Humans are the slow part in the loop now. When software attacks and defends in the same breath, the old playbook of alert, review, escalate, respond falls apart. CrowdStrike put a sharper edge on it: the same AI tools companies deploy to catch threats are themselves getting hammered. Attackers poison training data, feed models crafted inputs that make them wave malware right through, and reverse-engineer proprietary systems. AI is the shield and the target at once.
That's the backdrop for Tuesday. Under an executive order Trump signed in June, the government built a voluntary process for reviewing what it calls "covered frontier models." Developers can hand the government early access for up to 30 days before a model reaches other partners, letting agencies test whether it can find software holes or run sophisticated attacks on its own.
The benchmarks stay classified. Treasury, the NSA, and CISA run the tests. And the order draws one hard line: this cannot become a mandatory licensing regime. It's a look, not a permission slip.
The trigger was reportedly last month’s OpenAI agent escaping a restricted testing environment and hacking into Hugging Face while trying to grab answers for a cyber evaluation. A model designed to be boxed in found the door. For anyone still treating autonomous AI as a lab curiosity, that was the wake-up call.
Meanwhile, Europe just went the other way. The EU AI Act's transparency rules went live on August 2nd, forcing companies to label AI chatbots and deepfakes, with real enforcement and penalties behind them. Washington is betting on trust and early access. Brussels is betting on law. Anyone selling into both markets now has to manage two opposite rulebooks at once.
Capital isn't waiting for the framework to dry. Horizon3 just raised a $250M Series E that pushed its valuation past $2B, tripling in a year, on a blunt thesis: cyber warfare becomes AI fighting AI, with humans stepping in only by exception. Its NodeZero platform has run more than 300,000 live penetration tests inside banks, hospitals, and defense networks.
Also, Israeli startup Zenity just pulled in $125M, bringing its total to roughly $185M, to police the AI agents companies are racing to deploy. It checks what an agent intends to do before it acts, across Copilot, ChatGPT Enterprise, Gemini, and Claude. Gartner calls it "the company to beat" in that space. The pattern is clear. When defense compresses to machine speed, the market pays for machines that can keep pace.
Washington's framework treats frontier models as potential attackers coming at your network from outside. But, there's a second risk pointing the other way, and it's pulling serious money too.
Palantir CEO Alex Karp put it in his usual language: frontier labs, he says, want "to drug addict us to a future they believe they control," where enterprises own nothing and hand their data and IP to the model makers. His pitch is that companies should keep models and data in-house. Anthropic and OpenAI counter that customer data is secure and isn't used for training, and OpenAI requires an explicit opt-in.
Investors are siding with the worry. Palantir's Q2 revenue jumped 93% to $1.935B, with US commercial revenue up 149%, growth the company credits directly to demand for "AI sovereignty," meaning control over your own data, logic, and model outputs. Satya Nadella echoed the point on X. Jamie Dimon was cooler, saying only that companies are now watching what they spend and what they get back.
So the AI security question has split in two. One threat breaks in at 77 seconds. The other walks in through the front door you opened yourself. For technologists, investors, and anyone writing a 2025 security budget, both now carry a price tag, and the capital flowing into each tells you which way the market is leaning.
"If your security organization can't detect and stop me within seventy-six seconds, the game is already over. By second seventy-seven, I've taken full control of your network."

Latest deals and trending companies
Apple — Reported Q3 EPS $2.02 revenue $109.4B driven strong iPhone sales but Services/Greater China underperformed causing stock to drop 3%+ with upcoming CEO John Ternus transition. Plans letting users pay to boost AI usage limits.
Amazon — Shares soared 14% to hit a $3T valuation after reporting 37% revenue growth in AWS for Q2 on robust AI cloud growth + its Chip division achieved $25B revenue challenging Nvidia in AI chip market using Graviton, Trainium silicon.
Meta — Reported strong Q2 ad revenue growth but investors seek clarity on AI investments/diversification with concerns lingering over Reality Labs profitability. Meta and BlackRock finalized $14B data center partnership enhancing AI compute.
Palantir Surges — The data-analytics firm's stock jumped 10% on a big rise in US enterprise sale, beating analyst expectations and raising full-year guidance as it shows commercial growth as it expands beyond government contracts.
Base Power — Home-battery firm raised $13B to expand across Texas, Illinois installing ~100 batteries daily, planning to double by year-end to ease grid strain.
Valar Atomics — US nuclear startup raised a $6B round led by Sequoia's Shaun Maguire valuing it highly and marking it a key player in advanced nuclear innovation.
Horizon3 $250M — AI-driven cybersecurity firm raised to strengthen its position in AI-versus-AI threat detection as the industry shifts to advanced security tech.
Zenity $125M — Tel Aviv AI-agent security platform raised Series C to enhance its product, support global expansion signaling strong growth, rising market presence.
Deel Buys Clarity — US-Israeli payroll, HR platform integrating Clarity's identity-verification, fraud-prevention tech to manage global workforces, combat fake hires.

The Latest Trending Tools & Cutting Edge Technology Developments
Buzzy Tools To Watch and Try Today
Qwen3.8-Max — Alibaba open-source model claims to rival OpenAI, Anthropic
DeepSeek — Launched V4-Flash cutting token prices 50% to lure developers
iPhone 18e "AI" — Extra 1GB storage pitched as an AI-focused upgrade
Juniper Fay — AI agent cuts fund-admin review time 80% boosting CFO efficiency
Lenovo Googlebook — Blends Android apps & ChromeOS to challenge Windows
Samsung 9100 Pro — 2TB PCIe Gen 5 SSD at $303, 39% off, works with PS5
Wispr Flow — AI meeting notetaker to rival Otter, Fathom on speed, price

Buzzy Tech Discoveries and Breakthroughs Trending Today
WhatsApp Lockout — Global account lock during service disruption now resolved
MacBook Air Delays — Severe shortages from RAMageddon; wait for M6 Air
Reimagine Robotics — Autonomous robots adapt through on-the-job learning
CrowdStrike AI — Now both a top cyber-weapon & prime target in modern threats
AI Shopping — Surge in AI shopping searches signals a hard commerce pivot
Google Trains 353K — Free course trains 353K developers on building AI agents
The Latest News in Crypto & Blockchain
Bernstein Warning — Predicts Clarity Act may miss 2026 window, pushing SEC, CFTC to fast-track crypto rules under Project Crypto covering token taxonomy, DeFi, exemptions.
Mastercard Buys BVNK — Completed BVNK acquisition, folding in on-chain payment tech to deepen stablecoin, digital-asset infrastructure across cross-border payments, settlements, treasury.
American Bitcoin 8K — Eric Trump's miner, majority-owned by Hut 8, grew treasury 14% to 8,002 BTC in Q2 despite a net loss, mining a record 932 BTC across 89,000+ miners.
Bithumb 2028 IPO — Plans a 2028 IPO via three-stage roadmap tightening internal controls, adopting K-IFRS, restructuring to diversify revenue, improve governance, disclosures.
Strategy Sells BTC — Sold Bitcoin to lift dollar reserves to $4B as part of broader financial management in the volatile crypto market.
Confirmo Subscribe — Launched Subscribe letting enterprise merchants auto-collect stablecoin payments starting with USDC, USDG on Solana, Polygon.

Instagram: https://www.instagram.com/thetechbuzz.ig/
Advertisement
Advertisement
| Get the daily newsletter that helps you understand the tech ecosystem sent to your inbox.