What Telegram channels wrote about Data leak in August 2026: 8 posts from 8 channels, collected automatically from public Telegram channels.
⚡️10 сентября пройдет Всероссийская школьная олимпиада по информатике
Взлом сайта ФБР — 1 балл Слив базы данных Госдепа — 2 балла Доступ к камерам в туалете Джо Байдена — 3 балла Пароль Дональда Трампа от Госуслуг — 4 балла
We've Received Reports That Rippersec Is Harassing Pakistani Hacktivist Collectives and Allies Some Including With Doxing Threats Just because That particular Ideology and Targeting Didn't Match Rippersec's Ideology And Targeting
One Thing Is Criticizing The Other Thing Is Harassing a Collective Publicly To Isolate A Collective
Every Collective Has Some Questionable History Including Yours Rippersec
You're Group Publicly Say We Don't Target School And Hospitals ICS But CTI Reports Say You're Still Targeting Schools And Hospitals ICS Both In 2025-2026
Evidence: https://cyberasia.io/article/ddos/rippersec-launch-attack-against-network-school-due-ties-with-israeli
https://hackmanac.com/news/hack-tuesday-week-08-14-january-2025
https://cyberasia.io/article/scada/chat-control-protest-sauter-hmi-italy
These Are Latest Reports If We Go Back To 2024 Rippersec Did Data leak Against A Thai Hospital
https://x.com/MonThreat/status/1805965985913933960
Well RipperSec We Give you Choices
Either Your Group have to give independent evidence that this group did this or that
Or Stop accusing groups without independent evidence
#TeamAzrael
BEWARE: Crypto swap websites are tracking you - here's how they do it.
Most swap sites want you to connect your wallet first. That's when the data leak starts.
- device fingerprinting - wallet address logged - IP address tracked - transaction history linked
Pretty much everything about you is leaked at this point.
Unstoppable recently launched
swap[dot]unstoppable[dot]money
It collects nothing and allows you to swap without connecting your wallet. This doesn’t apply to all coins but most major ones are covered.
Also, all privacy coins are available including Monero, Zano, Zcash etc.
You get quotes from multiple providers, same risk labels as in the Unstoppable wallet app, and for the most part routes through DEXes.
⚡ Try Unstoppable Swap Bot !
@unstoppable_swap_bot
😎 Swap BTC, XMR, USDT, ZEC, USDT ...
• best rates • clear risk scores • no KYC (anonymous)
📉 На валидации 0.92. В проде 0.61.
Модель прекрасно работала в ноутбуке и развалилась, как только увидела настоящие данные. Утечка в признаках, дрейф, метрика, которая мерила не то.
Data Scientist: от данных до деплоя моделей — про весь путь после ноутбука.
⚙️ Данные и признаки: EDA, feature engineering, data leakage, стабильность признаков 📊 Обучение и оценка: baseline и gradient boosting, experiment tracking, error analysis, ложное улучшение качества 🚀 Прод: inference API, batch-инференс, quality gates, CI/CD, мониторинг data drift
▶️ 9 секций, 46 уроков. Первые уроки открыты — посмотреть можно бесплатно 🧩 Нужен Python и pandas; если просядете, тренажёры на платформе бесплатные
🏷 Внутри промокод ML_BOOKS — 30%, двое суток
➜ Открыть первый урок
📉 На валидации 0.92. В проде 0.61.
Модель прекрасно работала в ноутбуке и развалилась, как только увидела настоящие данные. Утечка в признаках, дрейф, метрика, которая мерила не то.
Data Scientist: от данных до деплоя моделей — про весь путь после ноутбука.
⚙️ Данные и признаки: EDA, feature engineering, data leakage, стабильность признаков 📊 Обучение и оценка: baseline и gradient boosting, experiment tracking, error analysis, ложное улучшение качества 🚀 Прод: inference API, batch-инференс, quality gates, CI/CD, мониторинг data drift
▶️ 9 секций, 46 уроков. Первые уроки открыты — посмотреть можно бесплатно 🧩 Нужен Python и pandas; если просядете, тренажёры на платформе бесплатные
🏷 Внутри промокод ML_BOOKS — 30%, двое суток
➜ Открыть первый урок
→ 100 Web Vulnerabilities, categorized into various types :
+ Injection Vulnerabilities: 1. SQL Injection (SQLi) 2. Cross-Site Scripting (XSS) 3. Cross-Site Request Forgery (CRF) 4. Remote Code Execution (RCE) 5. Command Injection 6. XML Injection 7. LDAP Injection 8. XPath Injection 9. HTML Injection 10. Server-Side Includes (SSI) Injection 11. OS Command Injection 12. Blind SQL Injection 13. Server-Side Template Injection (SSTI)
+ Broken Authentication and Session Management: 14. Session Fixation 15. Brute Force Attack 16. Session Hijacking 17. Password Cracking 18. Weak Password Storage 19. Insecure Authentication 20. Cookie Theft 21. Credential Reuse
+ Sensitive Data Exposure: 22. Inadequate Encryption 23. Insecure Direct Object References (IDOR) 24. Data Leakage 25. Unencrypted Data Storage 26. Missing Security Headers 27. Insecure File Handling
+ Security Misconfiguration: 28. Default Passwords 29. Directory Listing 30. Unprotected API Endpoints 31. Open Ports and Services 32. Improper Access Controls 33. Information Disclosure 34. Unpatched Software 35. Misconfigured CORS 36. HTTP Security Headers Misconfiguration
+ XML-Related Vulnerabilities: 37. XML External Entity (XXE) Injection 38. XML Entity Expansion (XEE) 39. XML Bomb
4 Broken Access Control: 40. Inadequate Authorization 41. Privilege Escalation 42. Insecure Direct Object References 43. Forceful Browsing 44. Missing Function-Level Access Control
+ Insecure Deserialization: 45. Remote Code Execution via Deserialization 46. Data Tampering 47. Object Injection 4 API Security Issues: 48. Insecure API Endpoints 49. API Key Exposure 50. Lack of Rate Limiting 51. Inadequate Input Validation
+ Insecure Communication: 52. Man-in-the-Middle (MITM) Attack 53. Insufficient Transport Layer Security 54. Insecure SSL/TLS Configuration 55. Insecure Communication Protocols
+ Client-Side Vulnerabilities: 56. DOM-based XSS 57. Insecure Cross-Origin Communication 58. Browser Cache Poisoning 59. Clickjacking 60. HTML5 Security Issues
+ Denial of Service (DoS): 61. Distributed Denial of Service (DoS) 62. Application Layer DoS 63. Resource Exhaustion 64. Slowloris Attack 65. XML Denial of Service
+ Other Web Vulnerabilities: 66. Server-Side Request Forgery (SSRF) 67. HTTP Parameter Pollution (HPP) 68. Insecure Redirects and Forwards 69. File Inclusion Vulnerabilities 70. Security Header Bypass 71. Clickjacking 72. Inadequate Session Timeout 73. Insufficient Logging and Monitoring 74. Business Logic Vulnerabilities 75. API Abuse
+ Mobile Web Vulnerabilities: 76. Insecure Data Storage on Mobile Devices 77. Insecure Data Transmission on Mobile Devices 78. Insecure Mobile API Endpoints 79. Mobile App Reverse Engineering
+ loT Web Vulnerabilities: 80. Insecure loT Device Management 81. Weak Authentication on loT Devices 82. loT Device Vulnerabilities
+ Web of Things (WoT) Vulnerabilities: 83. Unauthorized Access to Smart Homes 84. loT Data Privacy Issues 4 Authentication Bypass: 85. Insecure "Remember Me" Functionality 86. CAPTCHA Bypass
+ Server-Side Request Forgery (SSRF): 87. Blind SSR 88. Time-Based Blind SSRF
+ Content Spoofing: 89. MIME Sniffing 90. X-Content-Type-Options Bypass 91. Content Security Policy (CSP) Bypass
+ Business Logic Flaws: 92. Inconsistent Validation 93. Race Conditions 94. Order Processing Vulnerabilities 95. Price Manipulation 96. Account Enumeration 97. User-Based Flaws
+ Zero-Day Vulnerabilities: 98. Unknown Vulnerabilities 99. Unpatched Vulnerabilities 100. Day-Zero Exploits
Simplified Process: LLM generates responses → Humans evaluate → Preferred responses identified → Reward signal → Model optimized
The goal is to make the model more: Helpful, Safe, Aligned, Instruction-following
11. What is Model Alignment?
Model alignment means making an AI system behave consistently with intended human goals, values, and safety requirements.
An aligned model should: Follow legitimate instructions, Avoid harmful behavior, Provide useful responses, Respect safety constraints
12. What is Instruction Tuning?
Instruction tuning trains a model on examples containing instructions and desired responses.
Example: Instruction: "Summarize this article." → Expected Response: "Article summary..."
13. What is Supervised Fine-Tuning (SFT)?
Supervised Fine-Tuning trains a model using labeled examples.
Example dataset: Instruction → Expected Response: "Translate Hello" → "Bonjour"
14. What is Catastrophic Forgetting?
Catastrophic forgetting occurs when a model becomes better at a new task but loses some of its previous capabilities.
General LLM → Heavy Domain Fine-Tuning → Excellent domain performance → Reduced performance on some general tasks
15. What are the Risks of Fine-Tuning?
Overfitting, Bias amplification, Catastrophic forgetting, Poor-quality outputs, Data leakage, Privacy problems, High training costs
16. How do you prepare data for fine-tuning?
Raw Data → Cleaning → Deduplication → Filtering → Formatting → Train / Validation Split → Fine-Tuning
Good training data should be: Relevant, Accurate, Diverse, Consistent, High quality
17. How do you evaluate a fine-tuned model?
Compare the fine-tuned model against the base model.
Evaluate: Accuracy, Task completion, Response quality, Hallucination rate, Safety, Human preference, Domain-specific metrics
18. Fine-Tuning vs Prompt Engineering
Prompt Engineering: Changes instructions, Fast, Low cost, No training dataset required, Easy to iterate
Fine-Tuning: Changes model parameters, Takes training time, Higher cost, Requires training data, Good for specialized behavior
19. Fine-Tuning vs RAG
Use RAG when: Knowledge changes frequently, You need private documents, You need citations/grounding, You want to update knowledge without retraining
Use Fine-Tuning when: You need consistent behavior, You need a specific output style, You need task specialization
You can also combine them: Fine-Tuned LLM + RAG → Specialized + Grounded AI System
20. Interview Question: Design a Fine-Tuning Strategy
Strong Answer: "First, I would establish a baseline using the pretrained model and prompting. Then I would collect and clean high-quality domain-specific data, create train/validation/test splits, and determine whether full fine-tuning or PEFT such as LoRA is appropriate. I would fine-tune the model, evaluate it against the baseline, test for hallucinations and safety issues, and then deploy it with monitoring."
🎯 Key Interview Takeaways
Remember these five concepts:
Pretraining → General knowledge
Fine-Tuning → Specialized behavior
RAG → External/updated knowledge
LoRA/PEFT → Efficient model adaptation
RLHF → Human preference and alignment
These distinctions are extremely important in GenAI interviews.
Double Tap ❤️ For More
Decrypt covered our new report with CoinRabbit on financial privacy in crypto.
The report looks at both sides of the issue: illicit flows, scam compounds and laundering, alongside lawful privacy needs for civilians, companies, donors, journalists and high-value holders.
Public blockchains make balances and payment history readable to anyone. That can help investigators trace crime, but it can also expose users to stalking, extortion, corporate data leaks and physical targeting.
Our argument is simple: enforcement should focus on fiat off-ramps, exchange cooperation and AML controls where crypto meets regulated finance. Blanket surveillance of every wallet creates its own risks.
Privacy and accountability can be designed together.
Read the Decrypt coverage and full report 👇
