Chinese cyberespionage group Velvet Ant compromised an isolated critical infrastructure network in 2016 and remained undetected for a decade — by replacing Linux PAM and OpenSSH binaries with backdoored versions that embedded persistent access into the authentication process itself.
Meta is unwinding its $2 billion acquisition of Chinese-founded AI startup Manus following a divestiture order from Beijing citing national security concerns. The forced breakup highlights the mounting risks for Western companies pursuing Chinese AI assets.
Google filed suit in New York on June 12 against Outsider Enterprise, a China-based phishing-as-a-service operation that used Gemini to generate convincing fraud pages impersonating Google, USPS, E-ZPass, and banks. The FBI estimates the group stole 3.87 million credit card numbers and caused $1.9 billion in losses since July 2023. It is the first lawsuit Google has brought against threat actors for weaponizing its own AI.
EngineAI Robotics, the Shenzhen-based humanoid robot maker that opened a 12,000-square-metre factory in May and immediately began mass deliveries of its T800, has filed confidentially for a Hong Kong IPO. The three-year-old startup is valued at $1.5 billion after a $200 million Series B in April — and it is joining a stampede of Chinese robot makers racing to list while investor appetite for humanoid robots is at its highest point in history.
A Ukrainian drone developer has confirmed to New Scientist that ten AI-controlled quadcopters operating in fully autonomous 'terminator mode' killed Russian soldiers near Bakhmut in 2024 — with no human connection, no video feed, and no way to abort. Researchers say it is the most compelling evidence yet that an AI has independently made lethal decisions in combat.
The ShinyHunters group exploited a critical Oracle PeopleSoft zero-day for a full two weeks before Oracle published any patch, compromising more than 100 organizations and over 300 PeopleSoft instances worldwide. Universities bore the brunt of the attack, with the University of Nottingham among the confirmed victims and data on 455,000 individuals now circulating in leaked datasets.
The US government issued an export control directive on June 12, 2026, forcing Anthropic to disable its most advanced AI models worldwide. Anthropic is complying but publicly disputes that a narrow jailbreak technique justifies pulling a model deployed to hundreds of millions of users.
With H200 GPU shipments to China blocked by export controls, Nvidia is pitching its new Arm-based Vera CPU to Chinese cloud giants including Alibaba and ByteDance. Orders expected to ship August 2026.
A federal appeals court rejected Bankman-Fried's bid to overturn his FTX fraud conviction, calling him the main driver of 'one of the largest frauds on record.' His direct appeal path is now closed.
A global outage hit Meta's core platforms Thursday, locking out users across Facebook, Instagram, WhatsApp, Messenger, and Threads for several hours. Over 120,000 DownDetector reports filed at peak.
Microsoft's June 2026 Patch Tuesday is the largest in the company's monthly update history: 200 vulnerabilities patched, 38 rated critical, and six zero-days — three with exploit code already public. Researchers say AI-assisted bug hunting is why, and that this volume may not be a one-time event.
Anthropic launched Claude Fable 5 on June 9, its most capable publicly available model, alongside Claude Mythos 5 — the same underlying model with its security safeguards lifted, available only to vetted partners through Project Glasswing.