Questions & Alpha: Luan Wills on Why Hackers Target People and Not Code
Authored by James Newell, Foundation & Corporate Communications Director, GSR
Questions & Alpha: Luan Wills on Why Hackers Target People and Not Code
Authored by James Newell, Foundation & Corporate Communications Director, GSR
Include a disclaimer that this content is only for sophisticated investors.
Crypto settles in minutes, trades around the clock and usually has no intermediary to stop a payment once it leaves. Those are the strengths of the asset class. They also mean a single bad signature can cost tens of millions of dollars before anyone has time to react. And increasingly, attackers are going after people and keys rather than code.
In this edition of Questions & Alpha, GSR's interview series spotlighting our partners, clients, and the people building our business, Foundation & Corporate Communications Director James Newell sits down with Luan Wills, Head of Security, GSR, to discuss how the threat landscape has shifted over the past year, what good looks like for wallets and approvals, where clients most often get caught out, and why every security leader should already be planning for quantum computing.
In this conversation:
James Newell: Let's start with your role. What does it cover at GSR, and what does a typical week look like?
Luan Wills: My job is to protect GSR, our clients and their data, and to make sure we only move money when we mean to. That covers security engineering and operations, wallet and on-chain security, third-party risk, and governance. Increasingly it also covers how we adopt AI safely.
Alongside the regular checks that our controls are working, a lot of my week goes on new initiatives while they're still being designed.
James Newell: How is the threat landscape for a crypto market maker and OTC desk different from a traditional financial institution?
Luan Wills: The strengths of crypto are also what make exploits so costly. Settlement is fast and final, markets run around the clock, and anyone can use a protocol without going through an intermediary. Those are real advantages over traditional finance, but they also remove the pauses that give a bank time to stop a fraudulent payment.
Banks are hacked too. When attackers tried to move $951m out of Bangladesh Bank's account at the New York Fed in 2016, all but $81m was stopped along the way. On-chain, most assets have no one who can stop a transfer once it has been signed.
Stablecoins are the main exception. When Binance lost about $6.3bn in October 2026, Tether, Circle and other issuers froze around $3bn within a day, and Arbitrum's Security Council froze a further $30m on its network. But the attackers spread the funds across thousands of wallets within hours, and assets like ETH and BTC have no issuer that can freeze them. A freeze is also not a recovery: returning the funds depends on legal and governance processes that can take weeks or months.
So freezes help, but they rely on someone else acting quickly. The answer is to build safeguards that work at the same speed. A growing number of specialist firms in real-time monitoring, tracing, code audits and incident response now provide many of the checks that intermediaries give banks.
James Newell: What's the most significant shift in security risk you've seen in the last 12 months?
Luan Wills: The biggest shift is that DeFi is now being attacked the way centralised exchanges have been for years: through people, devices and keys rather than flaws in the code. Several of the largest DeFi losses in the past year started with a compromised signer or system.
In the first half of 2026, CertiK found that code vulnerabilities caused almost 60% of incidents but only about 12% of the money lost. Most losses came from compromised wallets and phishing.
Centralised exchanges still suffered the largest single thefts, including Binance's loss of about $6.3bn in October 2026, the largest on record. They also carry the widest contagion risk, because they serve a much broader public, many of whom have never used DeFi.
James Newell: What are the top three threats you spend the most time on?
Luan Wills: Three areas take most of my time.
Transaction integrity is crucial: every transaction we sign must do exactly what we approved, and nothing else.
The security of the DeFi protocols where we deploy capital goes well beyond their code. It also depends on who can upgrade or pause a protocol, who holds its admin keys and how they are protected, the oracles and bridges it relies on, and the protocols it is built on.
Social engineering is aimed at our clients as well as our staff. Attackers pose as recruiters, investors or business contacts to persuade people to install malware or approve a transaction. GSR is one of the oldest market makers in crypto, and that reputation makes our name useful to impersonators trying to reach our clients.
Third-party risk matters too, but it is well-trodden ground in traditional finance. The crypto-specific risks have fewer agreed standards, which is why they take up more of my time.
James Newell: How are attackers targeting people rather than systems, and how is AI changing both their toolkit and your defences?
Luan Wills: Fake hiring is the worst case: someone with a false identity joins the firm and gains legitimate access, which is why background screening matters. Most attacks, though, are phishing aimed at real employees, and AI has made it far more convincing. A deepfaked voice on a call or a fake conversation on Telegram is now quick and cheap to produce.
Impersonation on Telegram is a risk we have identified and are actively working to reduce. Anyone can check whether a Telegram contact is a genuine GSR employee at gsr.io/verify. We train all our employees on AI threats, including deepfakes, and we regularly spot and stop these attempts.
James Newell: What does the risk look like around custody, wallets and key management, and what does good look like for approvals?
Luan Wills: Some of the largest thefts, including Bybit in February 2025, came from signers approving a transaction that was not what it appeared to be. Each signer should verify what a transaction will actually do, using different hardware and software from the others, so that one compromise is unlikely to fool them all.
Multi-signature and MPC wallets split control of signing, but losses often come from what surrounds them: who can change the approval rules, restore a key from backup, or use the credentials and certificates that can start a transfer. Banks have applied dual control to this kind of access for decades, so that no one person can act alone and every action is recorded. Crypto sometimes forgets to apply that discipline.
James Newell: When market-participants are targeted, where do they most often get hit? And what mistakes do you see in how counterparties handle wallets, approvals and payment instructions?
Luan Wills: Address poisoning is one of the most common attacks clients face. The attacker creates an address whose first and last characters match one the victim regularly pays, then sends a tiny transfer from it so the lookalike hides in plain sight in their transaction history. Many wallets show only the start and end of an address, so the two look identical.
Changes to payment instructions should be confirmed through a separate, trusted channel, never the one the request came from. We will never ask clients for private keys, seed phrases or passwords.
James Newell: What risks are specific to token issuers, exchanges and institutions that are new to digital assets?
Luan Wills: For token issuers, the minting key is often a single point of failure. In March 2026, attackers used Resolv's minting key to create about 80 million unbacked tokens, because nothing in the contract capped how many could be issued.
Institutions new to the sector should not assume their custody platform is configured safely by default. Before the first transfer, they should check who can add a withdrawal address, change how transfers are approved, or use an API key on its own.
James Newell: If a client could only do five things this quarter to improve their security, what should they be?
Luan Wills: I'd pick three, not five.
Work from first principles. Identify the hazards your organisation faces, decide what must always be true to protect against them, and enforce it.
Run a live simulation of your worst-case scenario, with the people who would actually have to handle it.
Scan and test your systems continuously, using AI to find vulnerabilities faster. If you don't find them, someone else will.
James Newell: What are the early warning signs that something is wrong, and how should a team prepare for an incident?
Luan Wills: Early warning signs are usually small and scattered, so they need continuous monitoring rather than periodic checks. On-chain, that means watching for unusual transfers, upgrades to the smart contracts you depend on, and governance changes that weaken a control. Internally, it means changes to access, approval rules and payment details.
A response plan only helps if people have practised it and understand their role. Security, trading, operations, legal, compliance and communications all need to be in the room. Above all, there must be an executive with the authority to make the call, such as stopping outgoing transfers, without waiting for a meeting.
James Newell: What emerging risks should the industry be preparing for over the next two to three years?
Luan Wills: The emerging risk I'd focus on is quantum computing.
Token issuers who are not planning for it will struggle. The Ethereum Foundation has a good blueprint, with a dedicated post-quantum team since January 2026 and a roadmap targeting core protection by around 2029. But a network can only migrate itself, so issuers need to check what that leaves to them: their own contracts, admin keys and bridges.
More broadly, cryptography protects almost everything, from logins and payments to stored data, and encrypted data can be harvested now and decrypted later. Every organisation needs a full inventory of where it uses cryptography and a plan to move to post-quantum algorithms. NIST has already published the replacement standards and proposes retiring today's vulnerable algorithms between 2030 and 2035. Security leaders should be putting this in front of their decision makers now.
This material is provided by GSR (the “Firm”) solely for informational purposes. It is not intended to be advice or a recommendation to buy, sell or hold any investment mentioned. Investors should form their own views in relation to any proposed investment.
It is intended only for sophisticated, institutional investors and does not constitute an offer or commitment, a solicitation of an offer or commitment, or any advice or recommendation, to enter into or conclude any transaction (whether on the terms shown or otherwise), or to provide investment services in any state or country where such an offer or solicitation or provision would be illegal. The Firm is not and does not act as an advisor or fiduciary in providing this material.
This material is not an independent research report, and has not been prepared in accordance with any legal requirements by any regulator (including the FCA, FINRA or CFTC) designed to promote the independence of investment research.
This material is not independent of the Firm’s proprietary interests, which may conflict with the interests of any counterparty of the Firm. The Firm may trade investments discussed in this material for its own account, may trade contrary to the views expressed in this material, and may have positions in other related instruments. The Firm is not subject to any prohibition on dealing ahead of the dissemination of this material.
Information contained herein is based on sources considered to be reliable, but is not guaranteed to be accurate or complete. Any opinions or estimates expressed herein reflect a judgment made by the author(s) as of the date of publication, and are subject to change without notice. The Firm does not plan to update this information.
Trading and investing in digital assets involves significant risks including price volatility and illiquidity and may not be suitable for all investors. The Firm is not liable whatsoever for any direct or consequential loss arising from the use of this material. Copyright of this material belongs to GSR. Neither this material nor any copy thereof may be taken, reproduced or redistributed, directly or indirectly, without prior written permission of GSR.
Please see here for additional Regulatory Legal Notices relevant to US, UK and Singapore.