🔍 The Adidas Trionda - FIFA World Cup 2026's official ball: - Built-in motion sensor tracking position in real time - Works with 12 stadium cameras capturing data 50x/second - Powers VAR offside decisions via 3D player avatars - Has to be charged before every game The most advanced match ball ever made.

🧵 New Politico reporting on the Anthropic shutdown: - Andy Jassy flagged a jailbreak concern to the White House - By Friday it had reached Bessent, Cairncross & Lutnick - Then the two sides' stories completely diverge White House: Anthropic refused to cooperate for hours Anthropic: We got a 90-min deadline with no threat details.

🚨 Anthropic's response to the government is devastating: "If this standard was applied across the industry, it would essentially halt ALL new model deployments for ALL frontier model providers." Translation: 🔴 Fable 5 - shut down 🟢 GPT-5.5 - same capability, still running Antropic is the only one this was applied to. The company that reported the jailbreak? Amazon. Amazon's own AI? Still on. 👀

🤯 Andrej Karpathy, ex-OpenAI, ex-Tesla AI Director, now Anthropic researcher is reportedly unable to access Fable 5 & Mythos 5. Reason: he was born in Slovakia. 🇺🇸 The export control bars ALL foreign nationals 🏢 Including Anthropic's own staff 🤖 Including the people building the models

💰 Amazon invested $13B into Anthropic (expected to invest $33 Billion in all) 🔍 Amazon found a jailbreak in Fable 5. 📢 Amazon reported it to the Commerce Dept. 🔴 The government shut Fable 5 down. 🤖 Amazon also makes its own AI. It's called Nova. ⚔️ Nova competes directly with Fable 5. The "jailbreak"? Asking an AI to find code vulnerabilities. ✅ GPT-5.5 does the same thing. No jailbreak needed. ✅ GPT-5.5 is not shut down. ✅ GPT-5.5 has no export controls.

🚨 Amazon researchers told the US government they jailbroke Anthropic's Fable 5 & Mythos 5. Result? Commerce Secretary Howard Lutnick issued an export control directive shutting down worldwide access immediately. 🔒 Anthropic calls it a "misunderstanding" & is fighting back 👥 Even Anthropic's own foreign national employees are locked out

In 2015, Google bet $500M on SpaceX when it was worth $12 billion. Fast forward to 2026: 🚀 SpaceX IPO'd at $1.75T, the largest in history 📊 Google's stake: now worth ~$100 billion 💵 Return: 12,000% 🤝 Google also pays SpaceX $920M/month for AI compute

💀 151,900 tech workers laid off in 2026, and we are only halfway through the year. 📉 974 people losing jobs every single day 🤖 Nearly 50% of cuts explicitly blamed on AI 🇺🇸 US accounts for 80% of all global tech layoffs 📈 Same companies posting record profits & spending $700B on AI infrastructure The math is brutal.

🚨 U.S. government restrictions are now blocking Anthropic employees from accessing their own AI models, Fable 5 & Mythos 5. 🔒 A "huge percentage" of staff affected 🏛️ Government directive, not Anthropic's choice 🤖 The builders, locked out of their own creation

🚨 400 Arch Linux AUR packages compromised in a supply chain attack. 🎯 Attackers targeted orphaned packages & injected malicious build scripts 🔑 Stolen: browser passwords, SSH keys, API tokens, crypto wallets 🕵️ Malware hid itself as legitimate kernel threads to avoid detection If you use AUR, audit now.

💧 Amazon's data centers used 2.5B gallons of water last year. Water use per unit of energy in datacenters in world: 🟢 AWS: 0.12 L/kWh 🟡 Meta: 0.19 L/kWh 🔴 Microsoft: 0.27 L/kWh 🚨 Google: 1.15 L/kWh

🚨 CISA has added a new Cisco Catalyst SD-WAN zero-day to its Known Exploited Vulnerabilities catalog. CVE-2026-20245 (CVSS 7.8) — allows arbitrary command execution as root. But here's the dangerous part — attackers are chaining 3 CVEs together: 🔗 CVE-2026-20182 (CVSS 10.0) — auth bypass → front door 🔗 CVE-2022-20775 — privilege escalation 🔗 CVE-2026-20245 — root command execution Patch for CVE-20245 still in progress.

🚨 Google just sued a Chinese cybercrime network for weaponizing its own Gemini AI. The "Outsider Enterprise" used Gemini to: 🤖 Auto-generate phishing code 🌐 Deploy 9,000 fake websites 📱 Send 2.5M smishing texts in 2 weeks 💸 Defraud hundreds of thousands of victims FBI involved. RICO charges filed. First-ever AI abuse lawsuit.

⚽ Chinese-linked hackers have built 40 fake FIFA ticketing sites that: 👀 Track you through the entire checkout in real time 💳 Steal your card number, expiry & CVV as you type 📱 Intercept your bank's OTP verification code 💬 Use fake live chat to appear legit 80% of victims arrive via Facebook & Instagram.

🚨 ShinyHunters is claiming JCPenney breach and the deadline is June 15th. Allegedly stolen data includes: 📍 SSNs & dates of birth 💰 W-2 tax records & payroll files 🧑‍💼 Government-issued ID scans 📄 Driver's licenses

Checkpoint Security Report on Ransomware 🚨 Ransomware attacks jumped 48% year-over-year in May 2026. 📊 698 attacks globally in one month 🌍 Asia up 119% | EMEA up 40% | Americas up 39% 🏢 Business Services: 359% YoY 🛒 Consumer Goods: 223% YoY 🏭 Manufacturing: 50% YoY 61 active ransomware groups. The market is industrialized.

🤖 Claude Fable 5 just launched, and someone already claims to have jailbroken it. 👤 Hacker "Pliny the Liberator" says multi-agent prompting bypassed its safety layer 🔓 Claimed to extract info on cybersecurity, chemistry & explosives 🛡️ Anthropic disputes it and says it is NOT a real jailbreak Fable 5 was designed to fall back to a less capable model for high-risk topics.

⚽ The 2026 FIFA World Cup kicks off today, and so does the cyber threat season. 🌐 10,000 malicious World Cup domains already active 🎣 Phishing targeting fans AND event organizers 📄 Fake job sites stealing Google Workspace accounts 🏴‍☠️ State-linked actors could target critical infrastructure 🔎 FBI already warned of spoofed FIFA websites 5 million fans. 104 matches. And a massive attack surface for threat actors.

🚨 INTERPOL just took down Sniper Dz, a phishing-as-a-service platform active since 2015. 🌍 13 countries involved 👮 201 arrests made 🎣 45,000 victim records collected 🌐 20,000 phishing domains identified 🎯 Targeted PayPal, Netflix, Facebook, Instagram & more The platform was FREE to use thus lowering the bar for cybercriminals worldwide.

🇫🇷 France's encrypted government messenger, Tchap, got breached earlier this week. 💬 73,000 public sector employees affected 📧 Names, emails & org data exposed 📁 13.5GB of documents & media stolen 🔓 Public chat rooms were NOT encrypted 👤 Access gained via a compromised account