Deepfake

Seized Notebooks Revealed Sophisticated Scams
Hong Kong police uncovered a sophisticated fraud scheme that used artificial intelligence to deceive victims. The investigation led to the seizure of over HK$34 million (approximately USD 3.37 million). Notebooks confiscated by law enforcement revealed the criminals' methods, including the use of deepfake technology to appear more convincing.
How the Fraudsters Lured Their Victims
The fraudsters pretended to be wealthy single women, crafting stories about interests such as learning Japanese, playing golf, or tasting luxury wines worth over HK$100,000 (USD 12,850) per bottle. These methods were documented in the notebooks seized during the operation.
The investigation resulted in the arrest of 31 individuals connected to a criminal syndicate. This group used artificial intelligence to create realistic images of attractive women, which were then used to lure victims into romantic and investment scams.
The Problem of Deepfake Scams
Byron Boston, a former police officer and CEO of Crypto Track, warned that the combination of deepfake technology and social engineering presents significant challenges for investigators and law enforcement. AI-generated images make criminals more convincing and enable them to execute more complex scams.
Boston highlighted an incident from November 2022, where a fake video impersonating FTX founder Sam Bankman-Fried was used in a phishing attack targeting FTX users. This incident demonstrates how deepfake technologies can be exploited to steal cryptocurrency assets from victims.
Scams Targeting Young People
Confiscated materials revealed that the fraudsters specifically targeted young people seeking quick earnings. Victims were often convinced they were communicating with ideal women from Taiwan, Singapore, and Malaysia.
Challenges in Combating These Crimes
Boston emphasized that effective collaboration and swift action are key to fighting these sophisticated scams. However, he noted that many local law enforcement agencies, particularly in the U.S., lack the necessary tools and expertise to track stolen cryptocurrency or cooperate with international exchanges.
Criminals leveraging technologies like deepfake and social engineering remain a significant challenge for security forces worldwide.

#Deepfake , #CryptoFraud , #CryptoScams , #cybercrime , #CryptoNewss

Stay one step ahead – follow our profile and stay informed about everything important in the world of cryptocurrencies!
Notice:
,,The information and views presented in this article are intended solely for educational purposes and should not be taken as investment advice in any situation. The content of these pages should not be regarded as financial, investment, or any other form of advice. We caution that investing in cryptocurrencies can be risky and may lead to financial losses.“

Ngày 27/5/2025, Raj Gokal, đồng sáng lập Solana, trở thành nạn nhân của vụ rò rỉ dữ liệu cá nhân nghiêm trọng khi hình ảnh giấy tờ tùy thân và thông tin nhạy cảm bị phát tán trên Instagram, kèm yêu cầu tiền chuộc 40 Bitcoin (BTC), tương đương khoảng 4,36 triệu USD (CoinMarketCap, Bitcoin: 108,904 USD). Vụ việc không chỉ gây sốc cho cộng đồng crypto mà còn đặt ra câu hỏi lớn về bảo mật KYC và nguy cơ deepfake tạo bởi AI. Bài viết tóm tắt chi tiết sự cố, giả thuyết nguồn gốc, rủi ro bảo mật, và bài học cho nhà đầu tư.
Chi tiết vụ việc
Tài khoản Instagram của nhóm nhạc rap Migos (đã tan rã) bị xâm nhập và đăng tải hình ảnh nhạy cảm của Gokal, bao gồm hộ chiếu, bằng lái xe, và thông tin liên lạc như số điện thoại, email. Một phụ nữ, được cho là vợ Gokal, cũng xuất hiện trong ảnh với giấy tờ tùy thân. Kẻ tấn công đòi 40 $BTC , đe dọa tiếp tục phát tán nếu không được đáp ứng. Trước đó, Gokal đã cảnh báo trên X (20/5/2025) về các nỗ lực chiếm quyền email và mạng xã hội của ông, cho thấy đây là một cuộc tấn công có kế hoạch.
Vụ việc gây lo ngại về nguồn gốc dữ liệu bị rò rỉ, dẫn đến hai giả thuyết chính:
Lỗ hổng KYC: Hình ảnh có định dạng giống quy trình xác minh danh tính trên các sàn giao dịch crypto, nơi người dùng chụp ảnh cầm giấy tờ. Nếu là thật, dữ liệu có thể bị đánh cắp từ một sàn như Coinbase, vốn bị kiện trong tháng 5/2025 tại Mỹ (USA) vì vi phạm luật bảo mật sinh trắc học, liên quan đến việc chia sẻ dữ liệu KYC với bên thứ ba (Reuters).
#Deepfake AI: Hình ảnh có thể là sản phẩm giả mạo do AI tạo ra. Tháng 4/2025, podcaster Scott Melker tiết lộ kẻ lừa đảo dùng deepfake giả mạo bằng lái xe của ông, chiếm đoạt 4 triệu USD (Cointelegraph). Công nghệ AI hiện nay, như MidJourney, có thể tạo hình ảnh giả gần như hoàn hảo, khó phân biệt.
Rủi ro bảo mật trong hệ sinh thái crypto
Vụ việc của Gokal làm nổi bật các rủi ro bảo mật nghiêm trọng:
Lỗ hổng KYC: Theo CertiK (2024), 60% vụ hack crypto liên quan đến rò rỉ dữ liệu từ hệ thống KYC, gây thiệt hại 2,5 tỷ USD. Các sàn như Coinbase lưu trữ dữ liệu nhạy cảm, nhưng việc thuê ngoài xử lý dữ liệu làm tăng nguy cơ rò rỉ.
Deepfake tống tiền: Công nghệ AI tạo deepfake đã tăng 300% từ 2023–2024 (Chainalysis), được sử dụng trong lừa đảo và tống tiền. Các vụ như của Melker cho thấy khả năng thao túng niềm tin của nạn nhân.
Tấn công cá nhân: Các nhân vật nổi tiếng như Gokal là mục tiêu chính do giá trị tài sản lớn. Theo Glassnode, ví liên quan đến Solana nắm giữ 1,2 tỷ USD tài sản, thu hút tin tặc.
Tác động thị trường: Vụ việc làm giá Solana ($SOL ) giảm 5% trong 24 giờ, xuống 180 USD (CoinMarketCap), phản ánh tâm lý bất an.

Giải pháp quản lý rủi ro
Để giảm thiểu rủi ro, các biện pháp sau là cần thiết:
Cải thiện #KYC : Sàn giao dịch cần mã hóa dữ liệu và giảm chia sẻ với bên thứ ba. Binance sử dụng zero-knowledge proof để xác minh mà không lưu trữ dữ liệu nhạy cảm.
Phát hiện deepfake: Áp dụng công cụ AI như Sentinel để nhận diện hình ảnh giả mạo.
Bảo mật cá nhân: Sử dụng email và số điện thoại riêng cho tài khoản crypto, bật 2FA, và lưu trữ khóa riêng trong ví lạnh (Ledger).
Kiểm toán định kỳ: Các sàn cần được kiểm toán bởi Big 4 (PwC, Deloitte) để đảm bảo an ninh dữ liệu.
Tác động đến thị trường crypto
Vụ việc làm dấy lên lo ngại về bảo mật trong bối cảnh crypto tăng trưởng mạnh. Với 2,5 triệu người dùng tại Mỹ (Statista, 2025), các sàn như Coinbase phải đối mặt áp lực cải thiện bảo mật. Tuy nhiên, Solana vẫn có tiềm năng phục hồi, với TradingShot dự báo SOL đạt 250 USD vào Q4/2025 nhờ nâng cấp mạng. Vụ việc cũng thúc đẩy nhà đầu tư tìm đến các tài sản an toàn như Bitcoin và stablecoin (USDT, USDC).

Bài học cho nhà đầu tư
Bảo vệ thông tin cá nhân: Không tải giấy tờ tùy thân lên nền tảng không đáng tin cậy.
Chọn sàn uy tín: Ưu tiên sàn có kiểm toán Big 4 và bảo mật mạnh, như Binance.
Theo dõi rủi ro: Cập nhật tin tức hack qua CoinMarketCap, CertiK.
Đa dạng hóa danh mục: Kết hợp SOL, BTC, và stablecoin để giảm rủi ro biến động.
Gợi ý quản lý rủi ro
Sử dụng ví lạnh (Ledger, Trezor) để lưu trữ 80% tài sản.
Kiểm tra chính sách bảo mật của sàn qua báo cáo từ PwC hoặc Deloitte.
Theo dõi giá SOL và BTC trên CoinMarketCap, dự báo qua TradingShot.
Cảnh báo rủi ro
Đầu tư crypto, bao gồm Bitcoin và Solana, tiềm ẩn rủi ro cao do biến động giá và lỗ hổng bảo mật. Các vụ rò rỉ KYC hoặc deepfake có thể gây thiệt hại tài chính và danh tiếng. Nhà đầu tư cần bảo vệ dữ liệu trước hành vi sai trái như tống tiền và tự nghiên cứu kỹ lưỡng (DYOR). Thông tin trong bài chỉ mang tính tham khảo, không phải lời khuyên đầu tư.

Hong Kong Authorities Uncover Fraud Syndicate
Hong Kong authorities have uncovered a fraudulent syndicate that used deepfake technology to carry out a cryptocurrency scam worth $43 million. This scam resulted in the arrest of 27 individuals, including university-educated graduates.
According to local media reports, the syndicate had been operating since October of last year in an industrial unit spanning 4,000 square feet in Hung Hom.
Fraudsters Used AI to Create Attractive Female Personas
The fraudsters allegedly used artificial intelligence to swap their faces with those of attractive women during video calls. This deepfake technology allowed the male fraudsters to create appealing identities, helping them gain the trust of victims across Asia, including in Singapore, mainland China, Taiwan, and India.
Chief Superintendent Fang Chi-kin revealed that the group recruited digital media graduates to develop fake trading platforms and manage their online operations.
Victims Lured into Investing in Cryptocurrencies
Once trust was established, the fraudsters introduced cryptocurrency investment opportunities. To lure victims into investing, they showed fake trading records to convince them to invest large sums of money. Many victims realized they had been scammed only when they were unable to withdraw their funds.
Police Seized Equipment and Cash
During a raid on October 9, Hong Kong police seized more than 100 mobile phones, computer equipment, luxury watches, and over 200,000 Hong Kong dollars in cash. The suspects, aged between 21 and 34, face charges of conspiracy to defraud and possession of offensive weapons.
Authorities also discovered training materials instructing the fraudsters on effective tactics to manipulate victims. They found a "performance board" highlighting the top earners, with the most successful scammer allegedly earning $266,000 in a single month.
“They set up a performance board. Teams and members who successfully scammed the most victims were listed. The top earner last month made $266,000,” said Superintendent Iu Wing-kan.
Crypto Regulation in Hong Kong
These details of the crypto scam surface as Hong Kong tightens regulatory scrutiny of its cryptocurrency market. The Hong Kong Securities and Futures Commission (SFC) is currently reviewing nearly a dozen crypto platforms for potential licensing.
SFC CEO Julia Leung announced that 11 platforms applying for approval to operate crypto businesses had undergone on-site assessments. The SFC plans to issue licenses in batches to virtual asset trading platforms (VATP) to ensure regulatory compliance. Although these platforms are currently operating under a “deemed licensed” status, the SFC has warned traders not to engage with them until they are fully licensed.
The SFC’s approach includes granting conditional licenses to compliant platforms while stripping those that fail to meet regulatory requirements of their licensing qualifications.
#cybersecurity , #Cryptoscam , #hackers , #hacking , #deepfake

Stay one step ahead – follow our profile and stay informed about everything important in the world of cryptocurrencies!

Notice:
,,The information and views presented in this article are intended solely for educational purposes and should not be taken as investment advice in any situation. The content of these pages should not be regarded as financial, investment, or any other form of advice. We caution that investing in cryptocurrencies can be risky and may lead to financial losses.“

The crypto industry is constantly facing security issues. Therefore, it is important to remember to apply security methods. Ethereum founder Vitalik Buterin has repeatedly shared his wisdom with cryptocurrency users.
Recently, Buterin published an article analyzing the growing risks in the cryptocurrency sector, including the problem of “deep fakes” and their implications for security measures.
Let’s take a closer look at this article.
In his article, Buterin writes that every year it becomes harder and harder to recognize deep fakes as they become more realistic in appearance. He says that he recently became a target himself when a video featuring him was used to promote a scam and questionable investments.
He also emphasizes that audio and video recordings of a person are no longer a safe method of identifying their authenticity, citing the example of a company that lost $25 million due to a video conversation with a deep fake.
Cryptographic Signatures Are Not The Only Solution
Buterin criticizes the approach to cryptographic signatures as a method of verification. In his view, this approach ignores the broader context of security – the human factor.
Buterin argues that the practice of multiple signatures for transaction approval, which is intended to provide multi-level verification, can fail because an attacker can impersonate the manager not only for the last request but also for the previous stages of the approval process.
“The other signers accepting that you are you, if you sign with your key, kills the whole point: it turns the entire contract into a 1-of-1 multisig where someone needs to only grab control of your single key to steal the funds!” he notes.
Personal Questions As a Security Measures
Buterin writes: “Suppose that someone texts you claiming to be a particular person who is your friend. They are texting from an account you have never seen before, and they are claiming to have lost all of their devices. How do you determine if they are who they say they are?”
Probably inspired by Harry Potter, Buterin proposed a simple but powerful method of protection as a solution: “Ask them things that only they would know about their life.”
It is better to ask them, for example, about your experiences together:
When the two of us last saw each other, what restaurant did we eat at for dinner, and what food did you have?
Which movie did we recently watch that you did not like?
Which of our friends made that joke about an ancient politician?
The more unique your question, the better. Questions that make a person think, and they may even forget the answer, are good, but if your opponent claims to have forgotten, ask them a few more questions.
It’s always better to ask questions that relate to some “micro” details (what someone liked/disliked, personal jokes, etc.) than “macro” questions. Since the former are usually much harder for third parties to accidentally dig up (e.g. if even one person posted a photo of the dinner on Instagram, modern LLMs may well be fast enough to catch that and provide the location in real-time)
It Is Always Better to Combine Several Security Strategies
There is no perfect security strategy, so it’s best to combine several methods at once.
You can agree with a friend in advance on the passwords that you will use to authenticate each other. Or you can agree on a “duress” key, a word you can use to signal that you are being coerced or threatened.
The word should be common enough that you feel natural using it, but rare enough that you don’t use it accidentally in everyday conversation.
If you receive an ETH address, ask the person to send it to you through several communication channels (other social networks or messengers).
Protection against MitM attacks: Man-in-the-middle attacks are a common threat in digital communications. It involves an attacker covertly transmitting and potentially altering messages between two parties who believe they are communicating directly with each other.
To solve this problem, Buterin suggests using cryptographic protocols such as Transport Layer Security (TLS) and Secure Sockets Layer (SSL) to encrypt data in transit, making intercepted conversations unintelligible to outsiders.
Additionally, the implementation of end-to-end encryption in messengers ensures that only the users who are speaking can read messages, effectively eliminating the threat posed by these attacks.
The expert concludes the article with the following words: “Each person’s situation is unique, and so the kinds of unique shared information that you have with the people you might need to authenticate with differs for different people. It’s generally better to adapt the technique to the people, and not the people to the technique. A technique does not need to be perfect to work: the ideal approach is to stack together multiple techniques at the same time, and choose the techniques that work best for you.”
SoulBound Token
Vitalik Buterin is known for his brilliant ideas for projects. SoulBound Token was one of those projects. The Ethereum founder, along with lawyer Puja Alhaver and economist Eric Glenn, first proposed the concept in May 2022 to address some of the shortcomings of non-fungible tokens (NFTs) and similar ones.
SoulBound Token is an irreplaceable token valid for only one address that cannot be transferred or sold. This feature makes them ideal for representing assets that cannot be acquired through purchase, such as certificates of competence, reputation, medical records, etc.
SBT can be used for a variety of purposes, for example:
Maintaining medical recordsStorage of digital identity cardsMaintaining an employment record bookVerification of event attendanceAllows people to build a verified digital reputation based on past actions
Some companies and organizations have also used SBT to create a decentralized and secure digital identification system, for example:
Binance – launched its own SBT called Binance Account Bound (BAB) to improve Web3 identity verification and prevent fraud.
WhiteBIT – their Web3 service, WB Soul Ecosystem, allows for the recreation of a user’s identity in the Whitechain through the WB Soul and characterizes it according to your account.
Blockmate – discussing the use of SBT to display payment and debt history, enabling unsecured lending and improving credit scores.
Summary
Even though the expert recently stated that he is already “outdated” and will soon be replaced by a new talent, people still find his ideas and advice useful. They have repeatedly made life easier not only for cryptocurrency users but also for people not connected with crypto. And while Buterin hasn’t left cryptocurrencies yet, we will be waiting for new brilliant ideas from him.
#News #security #Buterin #deepfake

Crypto titan Changpeng "CZ" Zhao is sounding the alarm on a new wave of crypto scams. Fresh out of jail, he's urging the community to be on high alert for AI-generated deepfakes of him, spreading misinformation and promoting fraudulent crypto schemes.
"Fake me, real danger," CZ warns on X. Don't fall for the con!

OpenAI представила Voice Engine — ИИ, который синтезирует голос по 15-секундной записи.
Ты даёшь ИИ кусочек своей речи — он говорит как ты. И не просто копирует, а сохраняет интонации, эмоции, акценты. Пока модель тестируется в закрытом режиме — из-за риска фейков и обмана.
Где это уже применяют:
• локализация контента без актёров;
• голосовые помощники в медицине и обучении;
• озвучка YouTube и TikTok без затрат.
Но и угрозы серьёзные:
• дипфейки;
• мошенничество;
• «звонки от мамы» от ИИ.
Технология мощная. Осталось договориться, как её не просрать.
Где ты бы использовал свой голос-клон — в бизнесе, контенте или шутках?
#AI #OpenAI #VoiceTech #Deepfake
Подпишись — мы не просто говорим об ИИ, мы говорим его голосом.