Sophisticated attackers do not necessarily need to defeat an organisation's security perimeter if they can compromise a person, device or service that the organisation already trusts.  

The recent cyberattack on Binance powerfully illustrates this and reinforces a theme we discussed in our recent article on the NZSIS threat assessment: foreign-state activity, cyber threats, insider risk and rapidly developing technologies such as AI are increasingly interconnected enterprise risks.

For boards and senior management, the key questions are legal and governance ones: who holds critical authority, who bears the loss if it is misused, and whether the business could make the right decisions and meet its obligations in the first hours of an incident. We set out six questions below and discuss each in turn.

What happened?

On 29 October 2026, unauthorised access to an internal wallet-signing service at Binance was used to initiate about US$6.3 billion in cryptocurrency transactions. The attack has been attributed to a North Korean state-sponsored group. By 31 October, about US$2.85 billion had been frozen or intercepted, and Binance has said its Secure Asset Fund for Users (SAFU), an emergency reserve for customer losses, will cover any customer shortfall.

The incident is particularly relevant to New Zealand businesses in light of the New Zealand Security Intelligence Service’s (NZSIS) warning about the convergence of cyber intrusion, insider threats and foreign-state activity.

Insider risk is not limited to dishonest employees and can range from deliberate, influenced to unwitting activity. On the information currently available, the Binance attack did not involve anyone deliberately abusing their authority. Instead, a compromised employee device appears to have given the attacker a pathway to trusted internal systems.

Questions for boards and senior management

The incident raises six questions for boards and senior management, which we discuss in turn below.

  1. Do we know where our critical authorities sit, and do our delegations, approval rights and policies reflect how they are exercised in practice?

  2. If a trusted person, device or system is used to make an illegitimate transaction, who bears the loss, and is that outcome consistent with our fair conduct obligations and insurance cover?

  3. Do our cyber-security, insider-risk, operational-resilience, AML/CFT and sanctions frameworks connect with one another, or are these risks managed in separate silos?

  4. Could the right people take immediate action, and meet our notification obligations to regulators, agencies, counterparties and insurers, in the first hours of a serious incident, and have we tested how that would work?

  5. Do our contracts with technology and other key suppliers give us adequate security assurances, audit rights, incident notification obligations and recovery rights if a supplier is compromised?

  6. Does our AI governance address AI-enabled threats and AI-based monitoring, as well as our own use of AI?

1. Where do our critical authorities sit?

Some of a business’s most sensitive assets are authorities, such as the ability to approve payments, alter customer or payment details, change fraud-monitoring settings, or grant access to critical systems.

Binance's experience illustrates the point. Binance says the attacker was able to initiate transactions outside the normal withdrawal process using an internal wallet-signing service. A system may correctly recognise a credential that has authority to act, without the action itself being authorised by the business.

Identifying these authorities is a governance exercise. Board delegations, approval rights and internal policies should accurately reflect who can exercise these authorities in practice, and what checks apply. Directors' duties of care and diligence include overseeing the business's most significant risks, so boards should be satisfied they have enough visibility of how critical authorities are controlled, and what an attacker could do with them.

2. Who bears the loss?

Binance's ability to fall back on its SAFU fund shows that customer trust can be maintained by quickly showing how losses are dealt with, as well as by prevention in the first place. Who bears the loss when a trusted system produces an illegitimate transaction is a legal question best answered before an incident occurs.

For New Zealand businesses, the answer may depend on customer terms and conditions, fair conduct obligations under the conduct of financial institutions regime, contracts with service providers, and whether insurance would respond to a loss caused through trusted access. Boards should consider whether the business's current approach is one it would be comfortable explaining to customers and regulators, and whether the business can convey the position quickly and clearly so customers know where they stand in the event of an incident.

3. Are our risk frameworks connected?

The alleged North Korean involvement shows why different categories of risk should not be managed in isolation. An apparently technical cyber event can quickly raise issues of fraud, money laundering, sanctions, privacy, national security and regulatory compliance.

Boards should ask whether the teams responsible for each of these risks share information and escalate to one another, so that a single incident would be recognised as engaging each of them.

4. Can we act in the first hours?

The incident demonstrates the importance of speed. Attackers rapidly tried to convert freezable assets, while industry participants and law enforcement moved to trace and freeze them. Following any significant financial services incident, decisions about stopping transactions, contacting counterparties and engaging regulators may need to be made in minutes or hours.

Those decisions will often need to be made while notification obligations are running. A single incident could potentially engage suspicious activity reporting under the AML/CFT Act, sanctions obligations, notifiable privacy breach requirements under the Privacy Act 2020, and the incident notification expectations of the FMA or the Reserve Bank, each with its own trigger and timeframe.

An incident response plan that mainly addresses who receives a report the next morning may not be enough. Response arrangements should clearly set out who has authority to make decisions, escalation thresholds, regulatory and other notification timeframes, how legal privilege over investigations will be preserved, and how insurers will be notified in line with policy requirements. Those arrangements should be tested against realistic scenarios.

5. Are our supplier contracts fit for purpose?

Trusted access is not limited to employees. Technology and other service providers often hold privileged access to systems, data or payment processes. The security obligations, audit rights, incident notification requirements, liability provisions and insurance requirements in those contracts will be relevant both to preventing an incident and to determining who bears the loss if one occurs. The same issues arise in contracts with AI vendors.

6. Does our AI governance reflect the threat?

There is no basis to conclude that AI was responsible for the Binance compromise. However, generative AI makes impersonation, social engineering and the targeting of employees easier. AI is also part of the defence. Binance says it has expanded AI-based behavioural anomaly detection.

AI governance should therefore extend beyond an acceptable-use policy. It should address how AI is changing the threat environment, data use and accountability in contracts with AI vendors, and the use of AI-based monitoring tools, including privacy and employment obligations where employee behaviour is monitored.

How we can help

The legal issues raised by an incident like this arise before, during and after the event. We can help businesses to:

  • map critical authorities, delegations and approval rights, and test whether governance documents reflect how those authorities are actually exercised;

  • review customer terms and loss-allocation arrangements against any fair conduct obligations;

  • review incident response plans so that decision-making authority, escalation thresholds and reporting obligations under the AML/CFT, sanctions, privacy and financial markets regimes are clear, and contribute to board tabletop testing exercises;

  • review outsourcing, technology and other supplier contracts, including security obligations, audit rights, incident notification, liability and insurance requirements;

  • manage legal privilege during investigations, and coordinate engagement with regulators, law enforcement, insurers and counterparties if an incident occurs; and

  • develop AI governance frameworks that address both the use of AI within the business and the AI-enabled threats it faces.

Governance needs to extend beyond the perimeter

The Binance incident reinforces the NZSIS message that sophisticated external threats can become enterprise risks through the ordinary systems, relationships and people that organisations rely on. Resilience therefore depends on governance as much as technology. This means knowing where critical authority sits, who bears the loss if it is misused, and whether the organisation can make decisions and meet its legal obligations at the speed of a modern attack.

The lesson from Binance is ultimately a simple one. When an attacker can make themselves look like a trusted insider, controls designed only to keep outsiders out are no longer enough, and response readiness is critical.

Get in touch

If you would like to work through any of these six questions with your board or senior management, or have us review your incident response, notification or supplier arrangements, please get in touch with one of our experts below.

Contacts

Related Articles