How AI Detects Fraud in Banking

Posted on by Editorial Team

What is AI fraud detection for banking?

AI in fraud detection focuses on using ML technologies to reduce fraudulent activities in the banking and financial services sector.

By leveraging data, AI models are trained to distinguish between concerning activities and normal transactions, which aids financial institutions in mitigating the chances of fraud – detecting patterns far earlier than any human agent could be capable of spotting.

To enhance decision-making processes as well as risk and fraud management, AI solutions are being integrated into new and legacy workflows within financial institutions. ML algorithms powered by AI and trained on past data are capable of recognizing and blocking transactions deemed suspicious automatically. Furthermore, AI technologies may need human agents to validate confirmed suspicious transactions by completing additional safety checks. Furthermore, AI can also employ predictive analytics to forecast the types of transactions individuals may carry out in the future and identify if the new behaviors are anomalous.

AI finance technology (fintech) can aid in safeguarding against phishing scams, identity theft, payment fraud, credit card fraud and other forms of banking fraud on an individual level. These systems mitigate losses from such fraudulent activities. 

Customer experience can be impacted with AI systems due to false positives. Regardless of the way fraudsters choose to commit financial crimes, be it through unauthorized charges or even more illicit activities like money laundering, ensuring client accounts are secure alongside abiding to regulatory compliance is the primary focus of financial institutions.

Both fintech and other financial institutions are depending on AI as a fraud mitigation tool. With constant improvement, AI mitigation service providers and leading institutions expect thwarting fraud attempts will be abetted on an unprecedented level through automation.

How AI Is Implemented For Financial Fraud Detection

AI technology provides systems the ability to perform activities such as learn, adapt, problem solve, and automate with human-like intelligence. Even though AI technologies lack human-like cognitive abilities, when dealing with well defined systems, an AI that is trained on distinct tasks can operate much quicker and at a far greater scale than humans. 

Supervised and Unsupervised Learning 

AI systems put into action for preventing banking fraud are automated to attend to well defined activities. These AI models go through a process of supervised learning whereby they are fed large amounts of specially selected data which refines the model to perform tasks. This approach helps forge models that are able to detect requisite patterns for predetermined tasks.

On the other end, unsupervised learning enables drawing inferences from past data without given training documents.

Unsupervised learning  

The gaps of supervised training models using anomaly detection techniques can be filled with unsupervised learning anomaly detection techniques. With the help of these technologies, AI models are capable of identifying previously unpredicted but still abnormal behavior patterns. AI systems that incorporate unsupervised learning capabilities can sift through data to identify potential fraud long before human analysts would consider such actions a possibility.

Both supervised and unsupervised learning enable banks to automate the verification process. AI can scan the database for known fraud patterns and trigger alerts when new and unknown patterns that suggest fraud are detected.

Various Uses Of AI Technology  

Social Media Chatbot serves as a powerful example of an application using AI technology. It is one of the most widely used bots since it uses AI technologies as customer service agents and gives basic information as per the user query. 

In the realm of customer service, there are multiple other applications which the banking sector use to incorporate AI technology for detecting and preventing fraud: 

  • Real time systems: Automated AI programs are tasked with processing vast volumes of transactions and determining account activities within different parameters in real time identification and flagging of account suspicious activities which is sometimes referred to as intelligent automated systems. 
  • Help desk operations: With the use of advanced AI algorithms, traditional human operators tasked with proactive fraud detection can now talk to LLM-based AI assistants and use natural language so they can analyze complicated policy documents and large data sets.
  • Compliance enforcement: Financial institutions are facing enormous scrutiny to remain compliant with regulations. AI technologies assist banks with policy implementation by enforcing KYC compliance through automated ID checking for errors or fraud. These technologies also assist in the enforcement of Anti Money Laundering (AML) policies by identifying and flagging accounts, behaviors, and transactions such as the transfer of the same monetary value between unrelated accounts which are linked to money laundering schemes.
  • Fraud Detection: AI technologies are beneficial for applications that involve recognition of complex patterns for anomaly detection. There are some differences in AI systems known as graph neural networks (GNN) that specialize in handling data that can be modeled in the form of graphs, such as data common in the banking sector. GNNs are capable of processing billions of records and detecting patterns within vast data sets to track and capture even the most intricate fraudulent activities.
  • Risk Evaluation: AI and machine learning models are created using risk-based data with assigned weights to estimate chances of an event occurring. They also evaluate what action to take with the highest probable outcomes. In this regard, these models can make evaluations based on transaction amounts and history, frequency, location and behavioral tendencies, making them ideal for measuring risk. AI systems are capable of estimating risk associated with specific transactions as well as the exposure involved with issuing a loan or a line of credit to fraud-prone applicants.
  • Fraud Network Identification: Suspicious relationships between entities or clusters can be analyzed through machine learning techniques like graph analysis to identify fraud networks.

Differences between AI-Powered Fraud Detection and Traditional Methods  

AI technologies are transforming how fraud is detected and secured in banking, significantly enhancing efficiencies over older techniques. Where modern systems are leaps ahead of their predecessors, they trace some of their foundations from traditional models.  

Pros of Traditional Fraud Detection Systems  

Implementation simplicity: Traditional methods rely on heuristics, making them easier to execute. For example, automatically flagging any new transactions that exceed a certain predetermined and stagnant threshold based on the account’s historical data.  

Domain Intuition: An experienced fraud analyst brings useful domain knowledge and hunches to problem-solving. There are cases where only a traditionally trained person can adjudicate the validity of certain transactions or identify a fraud attempt.

Problems Encountered in Traditional Fraud Detection Systems

Scope Problem: Traditional fraud detection systems which use heuristics tend to be static and rely on fixed patterns (if X then Y relationships). While there is some merit to heuristics, in this case, the lack of efficiency results from ignoring the many interactions within a set of complex data.

Processed Transaction Problem: The ever-increasing volume of transactions from users leads to problems in systems that were built and manually controlled by a fraud detection expert. With a rising set of every minute of everyday systems, there is an increase in unattended data that needs to be processed. You could throw money at the problem, but hiring additional staff is inefficient, both cost-wise and output-wise.

High error rate: The traditional systems used rely heavily on rules which are arbitrary, leading to very low fraud detection rates. Because the rules are so harsh and agnostic to context, even ambiguous signals of potential fraud tend to over-trigger non-responsive systems. If an account configured for zero tolerance withdrawals tries to increase its drawdown by attempting to withdraw $200, which is not more than double the “allowed limit”, this will almost certainly trigger a blockage. Although such behavior is certainly unusual, it is only unusual from the perspective of a faulty rule-based system that attempts to operate under the guise of fraud detection. In reality, such actions cannot be labeled as “suspicious and unprecedented.” If anything, a customer will only want to make a large withdrawal instead of the normal one. The end result of all this is extreme low detection rates but massive resource waste in unproductive investigations.

Fraud Detection Using AI Technology: Benefits

Recognizing patterns better: AI technology uses sophisticated algorithms to process highly detailed and intricate data. When AI systems analyze data, they spot anomalies that would otherwise go unnoticed.

Unprecedented expansion: AI systems automated transaction monitoring far beyond human capabilities. Automated AI fraud detection systems offer transaction analysis and verification on-the-go, responding instantly unlike traditional systems.

Flexibility: Adaptable Algorithms Artificial Intelligence systems are trained to execute specific tasks. Their learning does not stop after training. An active AI algorithm retrains itself to improve techniques to intercept different forms of fraud as its systems keep working.

Drawbacks of AI-based fraud detection 

Fraud detection powered by AI systems has its downfalls. First off, the model needs adequate data to draw assumptions from. In order to successfully train an AI model, vast amounts of data suffices. The data needs to be collected, made through a thorough process (synthetic data), and finally filtered. The accuracy of an AI model is successful due to well trained data.

Tougher system applicability: Integrating AI systems into pre-existing structures could pose a challenge and become a burden to work with. Initially, these systems do seem to possess a high level of complexity and are hard to deal with, while in the end, prove to respond to positive ROI in long durations.

Applications of AI for fraud detection in the banking sector  

The adoption of AI-based fraud detection systems has proven beneficial for many banks and other financial institutions. LSTM based AI models, for example, improved American Express’s fraud detection systems by 6%. PayPal AI systems also enhanced their real-time fraud detection by 10% through round-the-clock global surveillance.  

In the banking sector, practical applications of AI for fraud detection are growing rapidly. Below are some of these applications.

Crypto tracing

The anonymity of cryptocurrency makes it an easy fraudster’s target. However, the sophisticated AI tools designed to combat fraud can monitor blockchains for abnormal behaviors such as unidirectional, streamlined fund transfers and trace misplaced or illicit payments. 

Verification chatbot

Bots equipped with AI can provide customer service and conduct verification processes as well. Phishing as well as identity theft because of an obvious tell in a particular interaction can easily be picked by chatbots, hence bots can be used to fish out scammers through language and user behavior analysis.  

Ecommerce fraud detection

To protect their clients from ecommerce fraud, banks can scan the client’s activities and purchase history and cross-check that with device information such as location to note unconventional transactions and block them from going through. Moreover, algorithms and purchase history can help identify dishonest ecommerce websites so that users can be warned in advance before making purchases through disreputable stores.

Problems Encountered with AI Fraud Detection Systems in Banking

AI fraud detection technology, an innovation in itself, is now actively and dramatically changing the banking industry. What’s more, there is even greater room for further improvement. However, with AI comes the challenges that this technology may bring.

Mistakes and AI ‘hallucinations’

We know that AI algorithms are improving their efficacy daily. But like every other technology, they are imperfect. An AI model is capable of generating ‘hallucinatory’ results, results that are false or inaccurate in nature. Within banks, the damages that can arise from such inaccuracies can be solved by creating very hyper-specialized models – models aimed at performing very specific tasks. However, such models also stifle the potential value that AI brings. Hallucinations, although not common, are highly prevalent, thus turning accuracy in AI banking fraud protection critical. 

Bias in Data Set

The issue of bias has persisted even before technology was involved with science, dating long back to the earliest days of data analysis. As much work as has been done to eliminate bias and discrimination from governing lending and account protection, the issue still remains. But just as critical, creating AI models by biased designers and engineers adds risks of discrimination, making the AI prone to disabilities based on gender, race, religion or even disability.

Compliance

The considerations concerning data privacy are crucial in the banking sector. AI models need considerable data which has to be collected and handled ethically. Compliance with data privacy regulations is equally critical when it comes to the application of AI. Indeed, the pace of technological development is extremely fast which means that lawmakers and regulators will have to revisit the question of whether our legal framework is suitable for the protection of customer privacy.

The Future of Blockchain Efficiency: Maximising Throughput Using Rollups

Posted on by Editorial Team

Author: Asutosh Mourya is Engineering Manager at Trili Tech – a blockchain research and development hub focused on the open-source Tezos blockchain.

Blockchain technologies are already part of our daily lives, even if we’re not consciously aware of them. They are applied in such areas as secure online transactions, supply chain management, identity verification, and even in creating and trading digital assets like non-fungible tokens (NFTs). Understanding how exactly blockchain works can shed light on its practical implications for enhancing efficiency and transparency in different aspects of our daily interactions.

In the article, I will share my views on how to maximise blockchain efficiency using rollups. Join me on this journey.

What is Blockchain

In terms of functionality, blockchain can be described as a series of data blocks linked in an uneditable digital chain. These blocks are stored in a decentralised environment, where each block’s information is verifiable by all participating computers. The decentralised structure ensures trust, validity, and usability, departing from traditional hierarchical systems.

Simply put you can think of a blockchain as a digital necklace made up of individual beads, with each bead representing a piece of information. These beads are linked together to form the necklace, just like blocks are linked to form the chain. Each bead holds details about transactions, who sent money and when it was sent.

In a blockchain, blocks not only contain transaction data but also a crucial element known as a hash. These cryptographic hash functions are fundamental to the blockchain’s operation. Hashes are represented by a unique series of characters like:

X23G9K1H4P8Q6L2V5

They act as the digital signature for a block, generated from its data. A key feature is that each block includes the hash of the previous block, forming a linked chain. This interconnectedness through hashes ensures the integrity of the blockchain network. Any attempt to modify the content within a block would alter the hash, signalling potential tampering to the network.

The incorporation of hashes creates a self-regulated network in the blockchain, eliminating the need for intermediaries. This design prevents third parties from monitoring or interfering with transactions, bolstering the security and reliability of the system.

What Makes Blockchain Special

Blockchains set themselves apart from other digital databases in distinct ways. Firstly, they operate as distributed databases, spreading data across multiple servers situated in various physical locations. This decentralisation enhances reliability, performance, and transparency compared to traditional databases. 

What is more, blockchains employ open-source software, allowing the entire network community to scrutinise the underlying code collaboratively. This transparency facilitates the detection and resolution of bugs, glitches, or flaws. 

Notably, once verified, new information can only be added to the blockchain; it cannot be altered. Security and trustworthiness are upheld by requiring majority consensus from network participants, promoting a shared responsibility model instead of relying on a single, central entity.

Current Limitations 

Though blockchain is getting widely adopted, it faces some challenges. One of them is scalability. In blockchain, it means handling more transactions without making things slower or less secure. To understand the grounds of this problem we need to get acquainted with the following concepts.

Block Time

This term refers to the average time it takes for a new block to be added to the blockchain. Different blockchain networks have varying block times. For example, Bitcoin has a block time of around 10 minutes, while Ethereum aims for a shorter block time of around 15 seconds. The challenge is to strike a balance: a shorter block time can lead to faster transactions, but it may also increase the likelihood of forks (divergent branches in the blockchain) and reduce security.

Transactions per Second

This metric indicates the number of transactions a blockchain can process in one second. The TPS varies widely among different blockchain networks. Traditional payment systems like Visa can handle thousands of transactions per second, whereas many blockchain networks, especially those focused on decentralisation and security, may have lower TPS.

For Bitcoin, the most used blockchain, the problem is its small 1MB block size, limiting the number of transactions and making fees higher. They suggested a solution called Segregated Witness (SegWit), but it’s not widely adopted yet. Ethereum, another big blockchain, has a 15-second block time, which is faster but also limits transactions per block. They plan to fix this with Ethereum 2.0, using Proof of Stake (PoS) and sharding. Other blockchains explore ideas like side chains or off-chain methods (like Lightning Network) to handle more transactions without slowing down the main blockchain. Solving scalability is crucial for making blockchains work better.

What are Rollups

Blockchain rollups are a Layer 2 solution for scaling cryptocurrencies, involving the consolidation of multiple transactions on a secondary blockchain (Layer 2). These transactions are then bundled into a unified piece of data, broadcasted onto the primary blockchain (Layer 1). 

In simpler terms, rollups extract transactions from the main blockchain, process them off-chain, compile them into a single data unit, and reintegrate them onto the primary chain. This process is why rollups are often referred to as ‘off-chain scaling solutions.’

At the most fundamental level, layer 1 scaling involves enhancing the scalability of the primary blockchain. On the other hand, layer 2 scaling entails relocating transactions from the main blockchain layer to a distinct layer that can interact with the primary chain.

Why Blockchain Rollups

Typical blockchain blocks have limited space, causing delays and increased costs as networks grow busier. Blockchain rollups solve this by consolidating transactions into one data piece off-chain, making processing more efficient.

How Do Blockchain Rollups Work?

Blockchain can store two types of information: transactions and data. While processing transactions on-chain is heavy, data resulting from transactions is lighter. Rollups merge transactions off-chain, submitting consolidated data to the mainnet, reducing the burden and enabling multiple transactions in one data piece. This enhances blockchain scalability.

A Step-by-Step Rollups Mechanics 

  1. Conducting Off-Chain Transactions

   – Engage in transactions directly on the rollup chain, acting as a blockchain platform.

   – Transaction processing occurs on the rollup chain, overseen by the “sequencer,” who validates, constructs L2 blocks, and submits transaction data with proofs to the primary L1 chain.

  1. Aggregating Batched Transactions

   – The sequencer organises multiple transactions into batches, collectively presenting them to the main L1 chain.

   – Batched transactions not only streamline processing but also reduce gas fees, providing a cost-effective experience for end-users.

  1. Ensuring On-Chain Security

   – Post-batching, the rollup chain delivers transaction data to a dedicated smart contract on the L1 chain.

   – Upon finalisation of the L1 block containing rollup transactions, the data becomes immutable, safeguarded against modification or censorship, ensuring constant data availability for verification.

  1. Generating Verifiable Proofs

   – Some rollups enhance transaction data with cryptographic “summaries” or “proofs.”

   – These proofs, serving as cryptographic assurances, are deposited on the L1 chain, validating the successful execution of the designated batch of transactions by the rollup.

Types of Blockchain Rollups

1. ZK Rollups (Zero-Knowledge)

 – ZK-SNARK: Uses short proofs for quick transaction processing and enhanced security but has vulnerabilities to certain hacks.

 – ZK-STARK: More scalable and transparent than ZK-SNARK, with larger proof sizes, providing improved security.

2. Optimistic Rollups

 – Assume all transactions are valid by default, approving them to the mainnet without extensive validation.

 – Use fraud-proving mechanisms to identify illegitimate transactions and penalise validators accordingly.

 – Depend on Ethereum mainnet for security, making them easier to implement and cost-effective compared to ZK-rollups.

Here are examples of operational rollup blockchains that simplify complex blockchain technology:

  • Optimism: an Ethereum optimistic rollup with a TVL of $700 million in November 2023. Optimism stands out for its standardised and open-source development stack, the OP Stack. Developers use it to launch their blockchains, and the native token is known as OP.
  • Base: An Ethereum optimistic rollup developed by Coinbase, one of the world’s largest crypto exchanges. Base does not have its own native token.
  • StarkNet: An Ethereum ZK-rollup leveraging zero-knowledge technology (STARK) for transaction computation and verification. The native token is called STRK.
  • Polygon Hermez: Polygon provides a suite of ZK-rollup solutions, Polygon Hermez is one of them, it employs Proof of Efficiency (PoE) consensus, allowing anyone to be a sequencer or aggregator.  Sequencers compile transactions, while aggregators validate and provide proofs to Ethereum. Polygon Hermez incentivizes honest behaviour and boasts decentralisation.

Benefits of Rollups

High Throughput

As can be seen from above, rollups, as a scaling solution for blockchains, deliver a remarkable boost in throughput. By efficiently processing and bundling multiple transactions together, they significantly enhance the overall capacity of the network. This surge in throughput ensures that a more extensive volume of transactions can be seamlessly handled, contributing to a smoother and more efficient blockchain experience.

Reduced Wait Time

Another prominent benefit ushered in by rollups is the substantial reduction in transaction wait times. Through the aggregation of transactions into batches, users experience quicker confirmation and processing.

Limitations of Rollups

Layer 2 blockchain solutions, while improving transaction speed and lowering costs, still face some important limitations. One concern is the risk of fraud by validators in Layer 2. Additionally, these solutions often sacrifice a bit of decentralisation for efficiency. Withdrawing from Layer 2 can be slow, as seen in plasma chains, and might involve added costs. 

Moreover, implementing Layer 2 solutions demands substantial computational power, making certain options less cost-effective for scenarios with lower activity. Despite these challenges, ongoing development in Layer 2 solutions remains vital for addressing blockchain scalability issues, playing an important role in the blockchain ecosystem’s future growth.

Hyperliquid: The Cost of Popularity – A Cautionary Tale for DeFi

Posted on by Editorial Team

A Complete Breakdown of the March 2025 Attack on Hyperliquid DEX

After its headline-making airdrop in November 2024, Hyperliquid quickly surged to the top ranks of decentralised exchanges (DEXs), overtaking heavyweights like Jupiter and dYdX. With blazing-fast transactions, zero KYC requirements, and deep liquidity pools, it became the go-to platform for crypto traders.

But in March 2025, the platform faced its biggest test yet.

On March 26, high-risk positions worth nearly $8 million were opened on Hyperliquid – positions that not only threatened the stability of the exchange but also put client assets in the Hyperliquidity Provider Vault (HLP) at risk. Behind the scenes, a series of suspicious trades and price manipulations hinted at a coordinated attack, leveraging vulnerabilities in both Hyperliquid and third-party platforms.

ExpertStack investigated the full timeline of events—analysing the attack, competitor responses, and the controversial decisions by Hyperliquid’s leadership that raised questions about decentralisation.

The Day of the Attack

Hyperliquid’s risk management relies on its HLP Treasury liquidity pools. Whenever a user opens a position, the system auto-executes a hedge. If a position is liquidated, the platform gradually buys back the asset—sometimes triggering a cascade.

On March 26, this mechanism was exploited to dramatic effect.

Attackers manipulated the price of JELLYJELLY, a low-liquidity token, on external platforms, creating a domino effect inside Hyperliquid’s system. At the time, HLP had around $290 million in its vault.

Key Timeline of Events:

⏱ Phase 0: “Market Preparation”
 From 10:50 to 12:15 UTC, JELLYJELLY price rose by 13%, then sharply crashed by 93%-a stress test before the main strike. The drop, from $0.1287 to $0.00831, aimed to liquidate long positions and destabilise the HLP.

⏱ Phase 1: Building a Delta-Neutral Position
 At 12:53 UTC, attackers opened large short positions ($4.08M) via address 0xde95…c91, while simultaneously placing long orders ($4.06M) through addresses 0x67fe…CA2 and 0x20e8…808 to hedge losses.

⏱ Phase 2: Triggering Liquidations
 Minutes later, they withdrew available margins and partially closed shorts to prompt liquidations. One short of ~$254K closed at $0.073978. Shortly after, a short position of nearly 400 million JELLYJELLY was dumped into HLP. The attackers swiftly moved ~2.76M USDC to Arbitrum – locking in a manipulated short price of $0.011282.

⏱ Phase 3: Final Blow – The Pump and Dump
 From 13:00 to 14:00 UTC, they aggressively bought JELLYJELLY across external exchanges, causing a 400% price surge to $0.05. Because Hyperliquid’s oracle relied on external spot prices, this manipulation immediately impacted derivatives and triggered massive unrealised losses.

JELLYJELLY price manipulation. Source: Lookonchain, Raydium.

Earlier transactions from March 15-25, flagged by Hyperliquid analysts, appeared to be dry runs – testing liquidation triggers and order types.

The Fallout

One attacker (a “whale”) successfully withdrew about $6.2 million, while another attempt to extract an additional ~$900K failed – leading to a net loss of ~$4K.

Hyperliquid halted trading, froze the price at $0.0095, and even ended with a slight gain of ~$700K. The team pledged full reimbursements to affected users.

Yet, this damage control sparked intense criticism. Influencers and Centralised Exchange (CEX) executives accused Hyperliquid of betraying the ideals of decentralisation and acting negligently.

Could CEXs Have Handled It Differently?

By April 8, 2025, Hyperliquid was handling ~$13B in daily perpetual trading volume – over 50% of the global DEX market share, per DeFi Llama. CoinGecko listed its open interest at $2.7B, beating major players like Deribit and even CEX arms like KuCoin and Crypto.com.

Hyperliquid accepts both USDC (via Arbitrum) and Bitcoin as collateral, giving users a rare ability to trade digital gold directly from their Web3 wallets.

On March 15, it captured 21% of Binance’s and 50% of Bybit’s volume in BTC futures – remarkable for a DEX.

But Hyperliquid’s team pointed fingers. Analysis suggested Bybit played a key role in the attack:

  • Oracle manipulation: Bybit’s spot data heavily influenced Hyperliquid’s margin price calculations.
  • Liquidity: Bybit’s deep order books allowed significant trades without slippage.
  • Market domination: With Binance not listing JELLYJELLY, Bybit’s pricing had outsized influence.

In short, attackers gamed Bybit’s data to distort Hyperliquid’s oracle.

Strategic Silence – or Not?

While manipulation happened through Bybit, other CEXs didn’t remain idle.

At 15:30 UTC, OKX listed JELLYJELLY perpetuals with 50x leverage. Binance followed at 16:00. The timing was… convenient.

Arthur Hayes, former BitMEX CEO, hinted at collusion. He cryptically framed OKX CEO Star Xu and ex-Binance chief CZ as having pounced on a weakened rival.

Hyperliquid’s Next Chapter

Despite the chaos, Hyperliquid’s core infrastructure remains strong. Built for scalability, it plans to integrate SVM and MoveVM, and leverages its custom HyperBFT consensus for potential L1 and L2 DeFi use cases.

In a March 31 interview with Wu Blockchain, LSD protocol developer Sean offered his perspective:

Binance and OKX feel threatened. They’re trying to recreate Solana-like dynamics by pushing meme tokens and onboarding users aggressively via BNB Chain.”

He also criticised influencers for disproportionately promoting CEXs and demonising Hyperliquid.

It’s fair to compete, but this narrative war is toxic. Centralised platforms aren’t saints – they have their own flaws. Their job is to enable fair trading, not run smear campaigns via influencers.”

Sean also acknowledged Hyperliquid’s shortcomings:

  • Closed-source limitations raise concerns about internal manipulation (e.g., MEV).
  • Insufficient blockchain transparency, such as lacking detailed account histories.
  • Dutch auction listings for low-cap tokens enabled overly large short positions.
  • Passive market maker design, if exploited, can lead to serious HLP losses.

The Bottom Line

The JELLYJELLY attack exposed real vulnerabilities – not just in Hyperliquid’s architecture, but in how DeFi interacts with centralised infrastructure.

More than just an isolated exploit, it was a wake-up call. It reignited debates around oracle reliance, systemic risk, and the blurry lines between CEXs and DEXs.

While Hyperliquid took swift corrective action and recovered financially, it now faces a trust battle—and a renewed mission to evolve its ecosystem without compromising its decentralised vision.

What’s next?
 That’s a question the entire crypto world should be asking.

Bernstein Praises Bitcoin’s Resilience Amid Market Turmoil

Posted on by Editorial Team

Bitcoin has shown remarkable resilience in the face of recent global market unrest triggered by U.S. President Donald Trump’s so-called “Liberation Day,” according to a new report from investment firm Bernstein, as cited by The Block.

In a time of widespread financial uncertainty, Bernstein analysts described Bitcoin’s performance as “simply impressive.” Unlike previous crises – such as the COVID-19 market crash, during which Bitcoin plunged between 50% and 70% – the current correction from its all-time high (ATH) has been a comparatively modest 26%.

This suggests that demand for Bitcoin is now coming from more stable, long-term capital,” the report noted.

While Bitcoin is traditionally viewed as a risk-on asset, Bernstein emphasised its evolving role as a long-term store of value. “We see Bitcoin as probabilistic gold over time – more volatile and liquid, but functionally similar,” analysts wrote.

Crypto analytics platform CryptoQuant provided additional context, stating that the current cycle’s 26.62% pullback is far less severe than the 83% drop in 2018 or the 73% decline in 2022.

Despite this relative stability, concerns remain. CryptoQuant recently highlighted increasing bearish conditions for Bitcoin, signalling caution in the near term.

Meanwhile, blockchain analytics firm Nansen predicted the crypto market could bottom out by June. CryptoQuant’s founder and CEO, Ki Young Ju, echoed this sentiment, suggesting that Bitcoin’s bull run has likely ended. He forecasts a period of decline or sideways trading over the next six to twelve months.

Crypto’s Quiet Takeover: How Lobbyists Infiltrated the White House

Posted on by Editorial Team

In the aftermath of the FTX collapse, the American crypto industry found itself at a crossroads. Disillusioned with the Democratic Party, many key players shifted their focus – and funding – toward Donald Trump’s inner circle. What we’re witnessing is the dawn of a new era: one where idealism takes a backseat to strategic alignment with traditional power structures.

At Expert Stack, we’ve unraveled the complex web linking major crypto entities to top officials in the new U.S. administration – and what they stand to gain.

The PayPal Powerhouse

Few figures are as influential in the American crypto lobbying landscape as Peter Thiel, PayPal co-founder and long-time venture capitalist. Thiel’s firm, Founders Fund, made a bold bet on Bitcoin back in 2014, exiting just before the 2022 crash with an estimated $1.8 billion in profit. By late 2023, the fund was back in the crypto game, investing $200 million ahead of the long-anticipated approval of Bitcoin ETFs.

Thiel’s influence doesn’t end with investments. His close associate and fellow venture capitalist, J.D. Vance – now the U.S. Vice President – benefited from a $15 million donation by Thiel during his 2022 Senate campaign. Vance later launched his own venture firm, Narya Capital, raising $93 million with backing from Thiel and Marc Andreessen, the co-founder of a16z.

Andreessen Horowitz (a16z) has been a key player in shaping crypto policy, funding the pro-crypto political action committee Fairshake with $140 million in support of Congressional candidates in 2024. Brian Quintenz, formerly a crypto policy director at a16z, was appointed chair of the Commodity Futures Trading Commission (CFTC) – a move that raised a few eyebrows in a now-familiar revolving door environment.

These players are closely tied to Elon Musk. In 2022, a16z, Binance, and Sequoia Capital were key investors in Musk’s Twitter acquisition. By 2024, they were advising Trump on key administration hires, per The New York Times. Notable figures include:

  • Marc Andreessen, a16z founder
  • Jared Birchall, Musk’s family office head and Dogecoin Foundation advisor
  • Sean Maguire, Sequoia Capital partner
  • Trey Stevens, Anduril co-founder
  • Shyam Sankar, Palantir CTO
  • David Marcus, Lightspark CEO and former Meta blockchain head

Musk himself leads the newly created Department of Government Effectiveness (DOGE), while David Sacks, former PayPal COO, was named the administration’s “crypto czar.”

But there’s another major player worth noting.

Tether’s Treasury Ties

One of the most strategic players in the crypto-political nexus is Tether, now among the top 20 buyers of U.S. government debt. The acquisition and custody of these bonds is managed by Cantor Fitzgerald, led by Commerce Secretary Howard Lutnick. The firm also reportedly owns 5% of Tether.

Digging deeper into Tether’s origins reveals billionaire Brock Pierce – co-founder of EOS, director of the Bitcoin Foundation, and an early pioneer in crypto’s grey areas, including the infamous Mt. Gox debacle. Pierce is only one degree removed from Trump via Steve Bannon, the former White House Chief Strategist. In the early 2000s, Pierce founded Internet Gaming Entertainment (IGE), a trailblazer in digital currency trading for online games. Bannon joined IGE as CEO in 2006.

Although Bannon later claimed that Pierce’s support for Trump in 2016 hindered their crypto collaboration, their shared past is telling.

In 2019, EOS purchased a domain from MicroStrategy for $30 million to launch the social platform Voice, with investors like Thiel, Bitmain, and Galaxy Digital’s Mike Novogratz backing the project.

A year later, Michael Saylor, CEO of MicroStrategy, made headlines by investing $250 million of the company’s capital in Bitcoin. By March 2021, institutional giants like BlackRock, Morgan Stanley, Vanguard, and Citadel held over 40% of MicroStrategy’s shares – suggesting broad financial sector approval.

Latecomers to the Power Game

More recently, firms like Coinbase, Grayscale, and its parent company Digital Currency Group (DCG) – led by Barry Silbert, a key early investor in Ripple, Coinbase, and CoinDesk- have entered the lobbying arena with renewed force.

DCG’s influence crosses party lines. Lawrence Summers, a former Treasury Secretary under Bill Clinton and Obama’s National Economic Council Director, serves as an advisor to the firm – highlighting bipartisan lobbying efforts.

The fall of FTX in 2022 dramatically shifted the political balance. The Democratic Party’s perceived mishandling of crypto regulation, personified by former SEC Chair Gary Gensler, alienated many in the industry – especially after revelations about his connections to Sam Bankman-Fried’s family.

As the landscape evolves, the new market consensus favours U.S.-based crypto ventures and dollar-backed stablecoins under more centralised oversight. Even previously neutral entities are now joining the political fray: Coinbase pledged $25 million to Fairshake for the 2026 midterms, while Ripple contributed $5 million to Trump’s inauguration fund.

Meanwhile, Saylor has taken a provocative stance, calling for the demise of gold as an asset class to weaken U.S. adversaries and tighten America’s grip on global capital reserves.

Conclusion

What began as a decentralised, anti-establishment movement has morphed into a sophisticated lobbying machine. Today’s crypto leaders are not challenging the state – they’re becoming part of it. And on social media, the revolutionary spirit of early crypto seems to be fading, replaced by a new wave of corporatised, state-aligned ambition.

How AI is Revolutionising Behavioural Biometrics For Authentication

Posted on by Editorial Team

Advanced technology like AI is completely changing the world of behavioural biometrics. A new and much better era of secure authentication procedures is being developed. For instance, monitoring the individual characteristic traits of a user like their writing tempo, movement of the mouse, and pressing on touchscreens makes possible AI powered behavioural biometrics and gives verification continuously and without interruption, thus making the experience better for the user and securing it more efficiently.

The History of Behavioural Biometrics

From the very beginning, every authentication system that utilises passwords and biometrics like fingerprints have been vulnerable to bypassing. On the other hand, behavioural biometrics employ dynamic interactions and actions performed by users which are nearly impossible to imitate, strengthening the security framework. Here, AI plays a critical role by examining massive amounts of user interactions in real-time by detecting small deviations that conventional systems would miss.

AI’s Role in Enhancing Behavioural Biometrics

With AI, relevant traits can be extracted from user interactions, and behavioural data may be turned into measurable metrics. User validation allows the application of intrinsic pattern recognition and the more advanced methods like neural networks and support vector machines. Continuous monitoring enables instant detection of anomalies, which aids in responding rapidly to security risks and unauthorised access.

Real-World Applications and Metrics

  1. Fraud Detection in Financial Services: A leading European insurer implemented behavioural biometrics to analyse how claimants interacted with online forms, detecting unnatural typing patterns and navigation behaviours indicative of fraud. This led to a 40% reduction in fraudulent claims within six months.
  2. Enhanced Customer Experience: An American health insurance company used behavioural biometrics for customer authentication, recognising users based on their interaction patterns with the company’s app. This approach reduced average customer service call times by 30%, significantly improving customer satisfaction.
  3. Risk Assessment Accuracy: A life insurance provider in Asia incorporated behavioural biometrics to refine risk assessment models by analysing lifestyle patterns affecting health and longevity. This led to more accurate premium calculations and personalised insurance packages.

Privacy and Needed Ethics

The application of AI in behavioural biometrics comes with notable ethical and data privacy issues. Even though these systems increase security, they need to be applied with care and responsibility, given the nature of the data. Security, user privacy, and inclusivity need to be balanced very carefully. Approaches like federated learning and edge computing provide the means for AI models to be trained on the user’s device, which greatly minimises the danger of breaches and strengthens compliance with privacy laws such as the GDPR.

Challenges and Future Outlook

Though promising, behavioural biometrics struggle with privacy and accuracy, as well as general user acceptance. Businesses in the field need to fortify protections and gain consent from users to avoid oversharing sensitive information. Usability and security have to be balanced because excessive false acceptances or rejections can undermine user trust in the system. Building trust requires addressing cultural differences along with the need for openness. Incorporating ethics focused on privacy, consent, and robust security makes the system more reliable.

With the advancement of technology, the integration of AI with behavioral biometrics will enhance authentication systems across numerous industries, providing users the ideal security and convenience.

Wallets of darknet marketplace Nemesis hit by US sanctions

Posted on by Editorial Team

The US Treasury Department’s Office of Foreign Assets Control (OFAC) has added 44 Bitcoin and five Monero addresses associated with the closed darknet marketplace Nemesis Market to the SDN.

The press release says they were controlled by Iranian citizen Behrouz Parsarad, who was allegedly the platform’s administrator.

On March 20, 2024, BKA seized Nemesis Market infrastructure in Germany and Lithuania, disrupting its operations. In the process, police confiscated digital assets worth €94,000.

The investigation began in October 2022.

The platform, created in 2021, sold drugs, stolen data and credit cards, as well as cybercriminal services, including ransomware, phishing, and DDoS.

Before the shutdown, Nemesis had an active audience of 30,000 users who carried out ~$30 million in drug transactions.

Parsarad received millions of dollars in commissions from the transactions and facilitated the laundering of digital assets, according to OFAC.

The administrator remains at large. According to the agency, Parsarad may have “discussed the creation of a new darknet market” with former suppliers.

Recall that in April 2022, German police confiscated the servers of the darknet marketplace Hydra and seized 543 BTC, and the US Treasury imposed sanctions on the platform.

That same month, an American court indicted Russian Dmitry Pavlov in absentia for administering Hydra, providing it with hosting services, conspiring to launder money, and distributing drugs. At the same time, the Meshchansky District Court of Moscow arrested Pavlov on another charge.

In December 2024, the Moscow Regional Court sentenced Hydra founder Stanislav Moiseev to life imprisonment and a fine of 4 million rubles.

The Billion-Dollar Heist: How Bybit Survived the Largest Crypto Hack in History

Posted on by Editorial Team

On February 21, the cryptocurrency world was shaken when Bybit, one of the largest Bitcoin exchanges, fell victim to a staggering $1.5 billion hack – marking it as the biggest cyber heist in crypto history. Despite the massive breach, the platform continued operating, thanks in part to swift crisis management and the backing of industry heavyweights.

How the Hack Unfolded

On February 21, on-chain detective ZachXBT reported suspicious ETH outflows from Bybit. We are talking about 499,395 ETH (about $1.46 billion at the time). The assumptions about the hack were confirmed by the company’s CEO Ben Zhou, and his employees almost immediately published a statement according to which the incident occurred when transferring ETH from cold multisig storage to a hot wallet.

The attackers replaced the transaction signing interface so that all participants in the procedure saw the correct address. At the same time, the logic of the smart contract was changed, and the hackers gained control of the ETH wallet and withdrew all the funds.

Zhou hastened to reassure clients and emphasized that the platform remains solvent and continues to process withdrawal requests, albeit with a delay: within about 10 hours after the hack, the exchange recorded a record number of withdrawal requests – more than 350,000. At that time, about 2,100 requests remained pending, while 99.994% of transactions were completed.

Nevertheless, the platform’s CEO still asked partners to provide a loan in ETH – the funds were needed to cover liquidity during the crisis period. As a result, more than 10 companies supported the exchange.

Huobi co-founder Du Jun contributed 10,000 ETH and promised not to withdraw it for a month. The co-founders of Conflux and Mask Network also announced the deposit of Ether to the exchange’s cold wallets. Coinbase Head of Product Conor Grogan wrote that Binance and Bitget sent >50,000 ETH there too.

According to reporter Colin Wu, 12,652 stETH (around $33.75 million) were transferred from MEXC to Bybit’s cold wallet.

The ETH price responded to the Bybit hack by falling to $2,625 (Binance), but recovered fairly quickly. By the evening of February 23, the quotes momentarily exceeded $2,850, after which they corrected to $2,690 (as of February 24).

Bybit representatives said that information about the incident has been “reported to the relevant authorities.” In addition, cooperation with on-chain analytics providers has allowed them to identify and isolate the associated addresses, limiting the attackers’ ability to “withdraw ETH through legitimate markets.”

As of February 24, Bybit has fully restored its Ethereum reserves (~444,870 ETH).

Who Was Behind the Attack?

According to ZachXBT, unknown individuals quickly exchanged some of the stolen mETH and stETH tokens for ETH via decentralized exchanges. 10,000 ETH were divided between 36 wallets.

The founder of DeFi Llama, 0xngmi, noted that the methods in this attack are similar to the incident with the Indian exchange WazirX in July 2024. At that time, Elliptic analysts concluded that North Korean hackers were behind the attack.

0xngmi’s assumption was confirmed by Arkham Intelligence. According to them, on the day of the Bybit hack, ZachXBT investigator “provided irrefutable evidence of Lazarus Group’s involvement in the hack”:

Its analysis contains a detailed analysis of test transactions and associated wallets used before the attack, as well as a number of graphs and timestamps. This data has been transferred to the exchange team to assist with the investigation.”

The founder of the AML service BitOK and crypto investor Dmitry Machikhin noted that the stolen cryptocurrency is actively being withdrawn from the Ethereum network to other blockchains. According to his observations, immediately after the hack, the assets were distributed to 48 different addresses.

At the second stage:

  • crypto assets from these addresses were gradually split into even smaller parts (50 ETH each);
  • funds were sent through bridges (eXch and Chainflip) to other networks.

The image shows how one of the 48 addresses splits the transactions into 50 ETH and goes to Chainflip.

According to Taproot Wizards co-founder Eric Wall, the North Korean hackers are likely to convert all ERC-20 tokens to ETH, then exchange the resulting ETH for BTC, and then gradually transfer the bitcoins to yuan through Asian exchanges. In his opinion, the process could take years.

ZachXBT reported that Lazarus transferred 5,000 ETH to a new address and began laundering the funds through the centralized mixer eXch, and then transferred them to bitcoin through Chainflip. The latter said that they have recorded attempts by the attackers to withdraw the stolen funds from Bybit in bitcoin through their platform. They disabled some front-end services, but it is impossible to completely stop the protocol, given its decentralized structure with 150 nodes.

The mETH Protocol team reported that they blocked the withdrawal of 15,000 cmETH (~$43.5 million) and redirected the assets from the attacker’s address to a recovery account. Tether CEO Paolo Ardoino said that the company froze 181,000 USDT related to the attack.

In a comment to ForkLog, Bitget CEO Gracie Chen emphasized that “the exchange’s systems have already blacklisted the attackers’ wallets.”

As of February 23, the attackers had exchanged 37,900 ETH (about $106 million) for bitcoin and other assets through Chainflip, THORChain, LiFi, DLN, and eXch. The hackers’ address still had 461,491 ETH of the 499,395 ETH stolen.

What to do?

After the hack, some community members started talking about rolling back the state of the Ethereum network to return the stolen funds. Thus, former BitMEX CEO Arthur Hayes noted that as an investor with large ETH reserves, he would support the community’s decision in the event of a chain rollback to an earlier state – as after the hack of The DAO in 2016.

Bitcoin maximalist Samson Mow also spoke out in support of restoring the blockchain, but leading Ethereum developer Tim Beiko criticized the idea. According to him, the Bybit incident involved an incorrect presentation of transaction data in the hacked interface, and not technical problems.

In addition, after the hack, the funds quickly spread across the complex ecosystem of the second-largest cryptocurrency by capitalization. “Rolling back” the network would mean canceling many legitimate transactions, some of which are related to actions outside the Ethereum network. The Vice President of Yuga Labs, nicknamed Quit, also drew attention to this. He added that many ordinary users would lose money, and the accounting systems of large players like Circle and Tether would collapse.

What’s the bottom line

The Bybit hack turned out to be the largest in the crypto industry so far. However, the head of Bitget did not find any reason to panic: according to her, the losses are equivalent to Bybit’s annual profit ($1.5 billion), and clients’ funds are completely safe.

The incident did not affect market sentiment either. According to Glassnode, the implied volatility of the first cryptocurrency is close to record lows. Price fluctuations against the backdrop of the hacker attack decreased after Strategy founder Michael Saylor published a chart of the company’s coin purchases.

This time, there was no platform crash or market panic, and a quick response and community participation helped restore liquidity and partially block the stolen assets. However, the incident highlighted a persistent problem – even large centralized platforms are still susceptible to attacks and vulnerable to hackers.

The Application of AI towards Real Time Fraud Detection on Digital Payments

Posted on by Editorial Team

The growth and development of the internet coupled with advanced digital communication systems has greatly transformed the global economy, especially in the area of commerce. Fraud attempts, on the other hand, have become more diverse and sophisticated over time, costing businesses and financial institutions millions of dollars each year. Fraudster activities and techniques have evolved from unsophisticated detection processes to contemporary automated methods based on rules through intelligent systems. Currently, artificial intelligence (AI) assists in both controlling and combating fraud, offering help to advance the sector of finance technology (fintech). In this article, we will explain the mechanics of AI in digital payments fraud detection focusing on the technical aspects, a real case, and relevant comments for mid-level AI engineers, product managers, and other professionals in fintech.

(um) let’s make tech work for us…

The Increased Importance of Identifying Fraud In Real-Time

The volume and complexity of digital payments, which include credit card transactions, P2P app payments, A2A payments, and others, continue to rise. Between 2023 and 2028, Juniper Research estimates that the cost of online payment fraud will climb beyond $362 billion globally. Automated and social engineering attacks exploit weaknesses such as stolen credentials and synthetic identities, often attacking within moments. Outdated methods of fraud detection that depend upon static rules (‘flag transactions over $10,000’) are ineffective against these fast paced threats. Systems are overloaded and angry customers worsen the problem, all the while undetected fraud continues to sail through.

Thanks to AI. Now, everything is seconds away, (we’ll repeat) all because of AI. With machine learning, deep learning and real-time data processing, AI can evaluate large amounts of data, recognize patterns, adapt to changes, and detect anomalies, all in a matter of milliseconds. For professionals in fintech, this movement is both a chance and a challenge: build systems that are accurate, fast, and scalable all while reducing customer friction.

How AI-Fueled Real-Time Fraud Detection Works

AI-enhanced fraud detection is supported by three tiers: data, algorithms, and real-time execution. Let’s simplify this concept for a mid-level AI engineering or product management team. 

The Underlying Information: For any front line fraud detection system, a payment transaction generated in real-time must be coupled with rich and high-quality data. This means diverse data, which includes transaction histories, user behavior profile data, device fingerprints, IP geolocation, and external sources such as chatter from the dark web. For instance, a transaction attempted from a new device located in a foreign country can be flagged as suspicious, when it is combined with a user’s base spending patterns. AI systems pull this data through streaming services such as Apache Kafka, or even cloud-native solutions like AWS Kinesis, which promises low latency. Data engineers must be willing to collect clean basic structured datasets, because the system performs poorly when the data given is of poor granularity. This is a proven lesson learned many times in the past twenty years for me.

Algorithms: The realm of AI has brought super advanced machine learning models into the world of detecting fraudulent activities, and these models are the backbone of AI fraud detection. Models with supervised learning capabilities work with labeled datasets (e.g. “fraud” vs. “legitimate”) and are proficient in recognizing established fraud patterns. Due to their accuracy and interpretability, Random Forests, and Gradient Boosting Machines (GBMs) are among the most popular models. Unfortunately, fraud is evolving much faster than data can be labeled and this is where unsupervised learning comes in. Clustering algorithms DBSCAN or autoencoders do not need previous examples and can pull unusual transactions for review. For example, even in the absence of historical fraud signatures, the sudden spike in small, rapid transfers can be flagged as it might indicate money laundering. Detection is further improved by deep learning models, such as recurrent neural networks (RNNs), that observe time series data (e.g. transaction timestamp) for hidden patterns and relationships.

Execution In Real-Time: Time is of the essence with digital payments. The payment systems must make a decision to approve, decline, or escalate a transaction in less than 100 milliseconds. This is only achievable by using distributed computing frameworks such as Apache Spark’s batch processing and Flink’s stream real-time analysis processing. Scaling inference is done using GPU-accelerated hardware, e.g., millions of transactions per second through NVIDIA CUDA, allowing for easy handling of over a thousand transactions every second. Product managers should remember that latency trade-offs can be detrimental when the complexity of the model increases; a simpler logistic regression may be suitable for low-risk scenarios, while high-precision cases require complex neural networks.

Real-World Case Study: PayPal’s AI-Driven Fraud Detection

To illustrate AI’s impact, consider PayPal, a fintech giant processing over 22 billion transactions annually. In the early 2010s, PayPal faced escalating payment fraud, including account takeovers and stolen card usage. Traditional rule-based systems flagged too many false positives, alienating users, while missing sophisticated attacks. By 2015, PayPal had fully embraced AI, integrating real-time ML models to combat fraud – a strategy we’ve seen replicated across the industry.

PayPal’s approach combines supervised and unsupervised learning. Supervised models analyze historical transaction data—device IDs, IP addresses, email patterns, and purchase amounts—to assign fraud probability scores. Unsupervised models detect anomalies, such as multiple login attempts from disparate locations or unusual order sizes (e.g., shipping dozens of items to one address with different cards). Real-time data feeds from user interactions and external sources (e.g., compromised credential lists) enhance these models’ accuracy.

Numbers: According to PayPal’s public reports and industry analyses, their AI system reduced fraud losses by 30% within two years of deployment, dropping fraud rates to below 0.32% of transaction volume—a benchmark in fintech. False positives fell by 25%, improving customer satisfaction, while chargeback rates declined by 15%. These gains stemmed from processing 80% of transactions in under 50 milliseconds, enabled by a hybrid cloud infrastructure and optimized ML pipelines. For AI engineers, PayPal’s use of ensemble models (combining decision trees and neural networks) offers a practical lesson in balancing precision and recall in high-stakes environments.

Technical Challenges and Solutions

Implementing AI for real-time fraud detection isn’t without hurdles. Here’s how to address them:

  • Data Privacy and Compliance: Regulations like GDPR and CCPA mandate strict data handling. Techniques like federated learning—training models locally on user devices – minimize exposure, while synthetic data generation (via GANs) augments training sets without compromising privacy.
  •  Model Drift: Fraud patterns shift, degrading model performance. Continuous retraining with online learning algorithms (e.g., stochastic gradient descent) keeps models current. Monitoring metrics like precision, recall, and F1-score ensures drift is caught early.
  •  Scalability: As transaction volumes grow, so must your system. Distributed architectures (e.g., Kubernetes clusters) and serverless computing (e.g., AWS Lambda) provide elastic scaling. Optimize inference with model pruning or quantization to reduce latency on commodity hardware.

The Future of AI in Fraud Detection

Whatever the future holds, it’s clear that AI’s role will only become more pronounced. For one, Generative AIs such as large language models (LLMs) could develop new methods of simulating fraud, while the involvement of blockchain technology could guarantee that the leger’s transaction records are safe from any possible modification. Identity verification through biometrics face detection and voice recognition will limit synthetic identity fraud.

As was noted previously, the speed, accuracy, and adaptability of AI in real-time fraud detection can enable users to effortlessly pinpoint and eliminate issues within digital payments that rule-based systems cannot alleviate. While PayPal’s success is evidence of this capability, the journey is not easy and requires fundamental discipline along with a well-planned approach. Now, for AI engineers, product managers, and fintech professionals, moving into this space is no longer purely a career change; it is an opportunity to build a safer financial system for all.