Modern Crypto Security Challenges in 2025: Rise of AI Fraud, Deepfake Attacks, Regional Risks, and the Role of Blockchain in the Confrontation

Blockchain and cryptography play a key role in countering modern, evolving artificial intelligence (AI)-based attacks, providing unique mechanisms for protecting and resilience systems in the face of increasing threat complexity.

The main ways blockchain and cryptography combat AI attacks are:

1. Decentralization as the basis for sustainability
   Thanks to the distributed architecture of the blockchain, data and control over the system are not concentrated in one point. Even if an attacker with AI tries to hack the system, he will only have access to a separate piece of data, and not to the entire database, which significantly complicates the attack and gives time for security services to respond. 1 4
2. Cryptographic protection of data and transactions
   The blockchain uses strong hashing and encryption, which makes data immutable and protects the integrity of information. This approach prevents counterfeiting and hacking using AI, since any changes to blocks require enormous computing resources and will be visible to the network. 1 6
3. Immutability and transparency
   Blockchain records cannot be changed retroactively, preventing fraudulent transactions and AI intervention without detection. Transparent transaction history allows for quick detection of anomalies and suspicious activity that automated AI may generate. 1 10
4. Cryptography in Access Management and Authentication
   Modern cryptographic methods provide multi-level authentication, including distributed key storage and multi-signatures, which prevents accounts from being compromised using AI tools – for example, during hacking or bypassing biometric authentication. 2 5
5. Integrating Blockchain and AI to Enhance Security
   Despite the initial incompatibility of the technologies, the combination of AI and blockchain allows for the creation of secure, decentralized AI models with controlled access rights, reducing the risks of counterfeiting and manipulation, and automating fraud detection using transparent and cryptographically protected data. 2 3 8
6. Consensus Mechanisms and Attack Defenses
   Consensus algorithms (Proof-of-Work, Proof-of-Stake, etc.) make a 51% attack extremely expensive and unlikely, ensuring the blockchain remains robust even when AI is used to coordinate hacking. 6
7. Using smart contracts with verification and adaptation
   Smart contracts can automatically check conditions and respond to suspicious activity, reducing the ability of AI to conduct stealth attacks and manipulations. 5

Thus, blockchain and cryptography create fundamental barriers to AI attacks through:

• decentralized data storage and management,
• strong cryptographic protection methods,
• transparency and immutability of records,
• multi-level authentication and access control,
• automated anomaly detection mechanisms.

These properties not only counteract AI attempts to hack or distort data, but also make it possible to increase the effectiveness of investigations and the speed of response to attacks. In the future, the development of AI and blockchain integration will allow the creation of more advanced security systems that adapt to new threats, taking into account their dynamism and scale.

Sources:
Current Blockchain Capabilities in Cybersecurity and Countering Cyberattacks (Antonyan E.A. et al.) 1
AI and Blockchain: Features of Interaction and Security (Habr) 2
AI and Blockchain: Joint Benefits and Challenges (Selecty) 3
Blockchain Technology in Ensuring Cybersecurity (Cyberleninka) 4
Modern Vulnerabilities and Features of Blockchain System Protection (TAdviser) 5
Blockchain Resistance to Attacks and the Role of Cryptography (ITSec.ru) 6
Prospects for Synergy between AI and Blockchain for Business and Security (Websoftshop.ru) 8

How AI is being used to help scammers bypass security systems

In 2025, fraudsters actively use artificial intelligence (AI) technologies to bypass security systems, making their schemes more sophisticated and difficult to detect. The main ways AI is used for criminal purposes include:

• Create convincing phishing messages . Generative language models help scammers automatically create realistic and well-formed emails that are difficult to distinguish from the real thing, significantly increasing the effectiveness of phishing attacks. 5
• Using deepfake technologies . With the help of AI, believable audio and video recordings are created that imitate the voices and faces of real people, such as relatives or bank employees. This allows you to deceive victims in video calls or phone conversations, inspiring trust and forcing them to transfer money. 1 3 5
• Bypassing biometric verification . Fraudsters use deepfake images and videos to fake identities when passing identification systems, including photos with documents or “face verification” on camera, which allows them to bypass the protection of banks and crypto exchanges. 2 5
• Automate and scale fraudulent operations : AI helps create scalable schemes, such as automated phishing campaigns and fraudulent attacks, allowing you to simultaneously reach a large number of victims with minimal costs. 5 8
• Jailbreak-as-a-service and AI model bypassing : Attackers use special techniques to bypass or “hack” AI systems to force models to generate malicious or prohibited content (such as ransomware codes), making it more difficult to monitor and combat abuse. 5
• Complex money laundering schemes . Using AI to build multi-layered schemes with multiple transactions through dozens and hundreds of crypto wallets, as well as cross-chain bridges and anonymous cryptocurrencies, making it difficult to track and investigate crimes. 5
• Doxxing and information gathering : Language models analyze big data and texts to help attackers extract personal information about victims, making it easier to create personalized scam scenarios. 5

Thus, AI increases the effectiveness and sophistication of fraudulent schemes, makes it more difficult to recognize and protect against them, and requires continuous development of protection technologies and increased user vigilance. 1 2 5

Why Deepfake Technologies Will Make Fraud More Dangerous in 2025

Deepfake technologies make fraud in 2025 significantly more dangerous mainly for the following reasons:

• High credibility and realism of fakes. Deepfake allows you to create audio and video recordings of voices and faces of people you know that are virtually indistinguishable from real ones. This creates strong trust in victims, who often do not doubt the authenticity of appeals from “relatives”, “friends” or “employees”, which makes it easier to deceive and steal funds. 1 2
• Availability and scalability of the technology. Previously, high-quality counterfeits required large resources and a professional team, but now any fraudster with a regular computer and the Internet can create deepfakes, which dramatically increases the number and variety of such attacks. 1
• Used for social engineering and psychological pressure. Fraudsters create fake video and audio messages with requests for urgent financial assistance or access to accounts. People tend to trust voices and faces, which makes the social factor key to the success of such attacks. 2 3
• Automation and complex fraud schemes. Artificial intelligence allows phishing campaigns to scale, turning deepfakes into a turnkey fraud tool, including the creation of synthetic identities and complex schemes for laundering stolen money. 5
• Deepfake attacks on financial and banking institutions are on the rise. In 2025, there will be a significant increase in such incidents, where fraudsters use deepfakes to deceive bank customers and employees posing as well-known company representatives or financial advisors. 2 5
• Abuse of identity verification and biometrics: Deepfakes allow fraudsters to trick identification systems, for example by creating synthetic facial images from different angles to bypass checks, making it more difficult to protect digital assets. 5

Thus, deepfake technologies enhance fraud by maximizing the effect of the human factor, increasing trust in deception, expanding the capabilities of attackers and complicating defense. The sharp increase in the use of deepfakes in fraudulent schemes has already been recognized as a serious threat to digital security in 2025. 1 2 5

How AI and Deepfake Technologies Will Improve Fraud Schemes in 2025

In 2025, artificial intelligence (AI) and deepfake technologies have greatly improved fraudulent schemes, making them more sophisticated and difficult to detect. 1 2 4 5

The main ways in which AI and deepfakes are used to improve fraud are:

• Creating believable deepfakes Fraudsters are actively using AI to generate realistic fake videos, audio recordings, and images, imitating the voices and appearance of familiar people. This allows them to convince victims to transfer money or provide sensitive data, imitating, for example, a call from a relative asking for help or a bank employee. The number of detected deepfakes increased fourfold from 2023 to 2024, accounting for 7% of all fraudulent attempts. 1 2 4 5 6
• Automating Phishing Campaigns Neural networks help scammers automate and scale phishing attacks, making phishing emails more realistic and increasing the number of potential victims. 2
• Synthetic ID Fraud Fraudsters combine real and fake information to create fictitious IDs. Banks use AI and deep learning to detect these fakes early in the onboarding process. 5
• Increasingly complex money laundering schemes Criminals use multi-layered schemes using dozens or hundreds of wallets, as well as anonymous cryptocurrencies (such as Monero) and cross-chain bridges, making it much more difficult to trace funds and conduct investigations. 6
• New methods of verification and identification Schemes have emerged where attackers ask the victim to “verify their identity” by filming their face from different angles, using this data to create deepfakes or bypass identification systems. 6
• Real Estate Fraud AI is being used to create fake photos of properties and listings from non-existent agents. In 2025, deepfake videos emerged of an agency “representative” giving a video tour of a non-existent property. 6
• Accessibility and Scalability of Fraud Technological advances have made fraud more accessible and scalable, eliminating the need for specialized skills. Fraud-as-a-Service (FaaS) platforms have emerged, offering ready-made tools for fraudulent operations. 1

Thus, AI and deepfakes allow fraudsters to create more convincing and automated deception schemes by exploiting the human factor and making their actions more difficult to detect. 2 6 7

1. https://humanconstanta.org/prognoziruemye-v-2025-godu-tendenczii-moshennichestva-s-pomoshhyu-ii/
2. https://www.tadviser.ru/index.php/%D0%A1%D1%82%D0%B0%D1%82%D1%8C%D1%8F:%D0%9C%D0%BE%D1%88%D0%B5%D0%BD%D0%BD%D0%B8%D1%87%D0%B5%D1%81%D1%82%D0%B2%D0%BE_%D1%81_%D0%B4%D0%B8%D0%BF%D1%84%D0%B5%D0%B9%D0%BA%D0%B0%D0%BC%D0%B8
3. https://www.cnews.ru/news/line/2025-07-23_sozdana_model_iikotoraya
4. https://cisoclub.ru/rossijskih-polzovatelej-predupredili-o-populjarnyh-scenarijah-moshennichestva-s-primeneniem-ii/
5. https://news.uppersetup.com/ru/banking/2025/05/16/financial-institutions-turn-to-ai-in-battle-against-rising-fraud/
6. https://www.sberbank.ru/ru/person/kibrary/articles/top-10-aktualnykh-skhem-moshennichestva
7. https://sodrujestvo.org/news/prognoz-populyarnyh-moshennicheskih-shem-v-2025-godu
8. https://togudv.ru/media/filer_public/f6/41/f641fe7c-75d2-4232-8925-2ca6cba065c5/kiber-2025.pdf
9. https://kredita.net/spravochnik/novye-skhemy-moshennichestva/
10. https://cyberleninka.ru/article/n/iskusstvennyy-intellekt-i-moshennichestvo-novye-ugrozy-i-sposoby-protivodeystviya

Sources:
SecurityLab — Deepfake Fraud: How to Protect Yourself from Counterfeiting (2025) 1
T
TAdviser — Deepfake Fraud (2025)
LegalAcademy — Deepfakes and Liability (2024) 3
Anti-Malware.ru — Deepfake Attack Forecasts 2025 5

1. https://www.securitylab.ru/blog/personal/xiaomite-journal/355873.php
2. https://www.tadviser.ru/index.php/%D0%A1%D1%82%D0%B0%D1%82%D1%8C%D1%8F:%D0%9C%D0%BE%D1%88%D0%B5%D0%BD%D0%BD%D0%B8%D1%87%D0%B5%D1%81%D1%82%D0%B2%D0%BE_%D1%81_%D0%B4%D0%B8%D0%BF%D1%84%D0%B5%D0%B9%D0%BA%D0%B0%D0%BC%D0%B8
3. https://legalacademy.ru/sphere/post/dipfeiki-i-otvetstvennost-za-ih-rasprostranenie-chto-vazhno-znat
4. https://dialog.info/s-nachala-2025-goda-obnaruzheno-67-dipfejkov-ot-ih-obshhego-chisla-za-proshlyj-god/
5. https://www.anti-malware.ru/analytics/Threats_Analysis/Waiting-for-Deepfake-attacks
6. https://ralj.ru/article/download/(2025)1.8/14647
7. https://www.forbes.ru/tekhnologii/539953-eksperty-soobsili-o-rezkom-roste-cisla-mosennicestv-s-ispol-zovaniem-dipfejkov
8. https://sbersova.ru/sections/protection/polnaya-kopiya-kak-raspoznat-dipfejk
9. https://spb.bankstoday.net/last-articles/deepfake-ugrozy-kak-zaschitit-sebya-i-svoi-biometricheskie-dannye-v-2025-godu
10. https://www.gazetametro.ru/articles/88-millionov-zhertv-kak-ne-popastsja-na-udochku-aferista-26-06-2025

1. https://vc.ru/ai/1752843-kak-ii-zashishaet-ot-fishinga-telefonnyh-moshennikov-i-aferistov-na-marketpleisah
2. https://trends.rbc.ru/trends/industry/67a05f319a79471a124f1732
3. https://chataibot.ru/blog/neyroset-i-moshenniki/
4. https://cisoclub.ru/fishing-novogo-pokolenija-kak-zloumyshlenniki-prevrashhajut-ii-v-oruzhie/
5. https://hot-digital.ru/2024/05/5-ways-criminals-are-using-ai/
6. https://www.unite.ai/ru/how-scammers-use-ai-in-banking-fraud/
7. https://botfaqtor.ru/blog/kak-ii-boretsya-so-sklikivaniem-i-sohranyaet-roi/
8. https://habr.com/ru/articles/855248/
9. https://companies.rbc.ru/news/rt6UuPqeY7/iskusstvennyij-intellekt-pomogaet-bankam-borotsya-s-moshennichestvom/
10. https://www.itsjournal.ru/articles/special-report/iskusstvennyy-intellekt-zashchitit-i-zashchititsya/

1. https://aprp.msal.ru/jour/article/viewFile/1464/1218
2. https://habr.com/ru/companies/yandex_praktikum/articles/753836/
3. https://selecty.ru/blockchain-al
4. https://cyberleninka.ru/article/n/blokcheyn-tehnologiya-v-obespechenii-kiberbezopasnosti-gosudarstva
5. https://www.tadviser.ru/index.php/%D0%A1%D1%82%D0%B0%D1%82%D1%8C%D1%8F:%D0%91%D0%BB%D0%BE%D0%BA%D1%87%D0%B5%D0%B9%D0%BD_(Blockchain)
6. https://www.itsec.ru/articles/ataka-51-i-ustojchivost-blokchejna-bitkoina
7. https://ib-bank.ru/bisjournal/news/9037
8. https://www.websoftshop.ru/information/articles/blockchain_technology/blockchein_and_ai/
9. https://www.gate.com/ru/learn/articles/how-will-ai-revolutionize-ethereum/2134
10. https://cyberleninka.ru/article/n/analiz-blokcheyn-tehnologii-osnovy-arhitektury-primery-ispolzovaniya-perspektivy-razvitiya-problemy-i-nedostatki