Data breaches · Liability

When data breaches break platforms: who carries the loss?

By V Kumar Sharma · 19 November 2026 · Based on my keynote "Defences Breached, Platform Compromised" (April 2026)

In short
  • Many platform losses start with someone else's breach: a vendor, an API integration or a telco.
  • Once a platform is on notice of a breach, the question becomes whether it kept users secure against foreseeable threats, including ones arising from that breach.
  • Platforms can shift risk to users through forced resets, re-verification and stronger authentication, but every step costs user trust.
  • After a loss, look at every party in the chain, including vendors, off-ramps, banks and insurers, not just the threat actor.

How attackers turn leaked data into losses

Threat actors rarely break the front door of a well-run platform. They come in sideways, using data leaked from somewhere else:

  • Phishing and business email compromise. The attacker sits between a company and its counterparty and redirects a payment for goods or services to an account they control.
  • API key leaks. When a third-party tool holding users' exchange API keys is breached, the attacker can trade on those accounts without ever logging in. In the 3Commas incident, around 100,000 API keys were reportedly leaked.
  • Contra trading. With access to many compromised accounts, the attacker opens and closes offsetting positions, moving value from victims to accounts they control. In some incidents withdrawals were halted on the "primary" compromised account, but trading was not, so the drain continued.
  • Telco metadata. Call logs between executives can reveal an organisation's hierarchy and its key targets. Network data can reveal service account names and DNS configuration, which show exactly where to get into a network.

The platform's duty: what was it securing against?

When a breach originates outside the platform, I ask a sequence of questions:

  1. Was the platform put on notice of the breach or the vulnerability?
  2. Does it owe users an ongoing obligation to keep their assets and the platform secure?
  3. Does that obligation extend to foreseeable threats, including API integrations and third-party data breaches?
  4. If not, what exactly is the platform securing against?

Post-incident reviews often focus narrowly on software, security controls, configurations and system logs. They can miss organisational, policy and management failures, such as a database left without a password, or questions about who knew where to look. In a dispute, those are often the decisive facts.

The platform's dilemma: shifting risk without destroying trust

After a third-party breach, platforms have tools to shift risk onto users. Each one carries a cost:

  • Mandatory credential changes (new passwords or even new email addresses), with account limits until users comply. This breaks the attacker's access path and makes the leaked data obsolete, but causes real friction and reputational damage. If a user ignores the instruction and is later compromised, a court or tribunal is more likely to find that responsibility for the loss has shifted to the user.
  • Authentication upgrades. Mandatory 2FA/MFA is an industry standard and a reasonable security enhancement. Biometrics give strong verification but raise privacy concerns. Behavioural analytics work invisibly in the background.

Do too little and users conclude the platform doesn't take security seriously. Do too much and users feel punished and leave. The right answer is usually a transparent, proportionate response with user choice where possible.

The cascade effect

If a platform does nothing about a third-party breach, the damage tends to cascade:

  1. Data exfiltration: credentials, personal data, transaction data
  2. Account takeovers and unauthorised transactions
  3. Loss of user trust and users leaving the platform
  4. Regulatory investigation, enforcement and litigation
  5. A threat to business continuity

The speed of the response decides which stage you can stop it at.

After the loss: who else is in the chain?

Recovery starts by identifying every party involved, looking both forward and back along the flow of data and funds:

  • Third-party vendors where the breach originated, and the contracts that govern them
  • Parties with monitoring obligations, such as those obliged to monitor transactions or keep the platform secure
  • Off-ramps and recipients, with possible claims in knowing receipt or dishonest assistance
  • Banks, where a Quincecare-type duty of care may be relevant
  • Insurers: what does the policy cover, and has it been triggered?

Negligence has been used successfully to allocate liability in these cases, and contributory negligence by the user will often be argued in response.

Frequently asked questions

Is a platform liable for losses caused by a third-party data breach?
It depends on the facts, but key questions are whether the platform was on notice and whether it took reasonable steps against foreseeable threats arising from the breach.
What is contra trading?
Using several compromised accounts to place offsetting trades that move value from victims to the attacker. It can continue after withdrawals are halted, unless trading on affected accounts is suspended too.

This article is general information as at the date above and is not legal advice. It does not describe any particular client matter.

V Kumar Sharma

V Kumar Sharma is a Singapore disputes lawyer focusing on cyber incident response, digital asset recovery and technology disputes. ALB Asia 40 Under 40 2026.