Blog 19.11.2026

Third-party code, your responsibility

Man typing at a kitchen table looking at a laptop.

A tool designed to improve the customer experience can also create an entry point for an attacker. According to security company Report URI, an attack on Optimizely on 16 November 2026 delivered malicious JavaScript to visitors of online services operated by 178 organisations, including Air France-KLM, IKEA and GitLab. The code enabled the theft of cryptocurrency and payment card details, as well as attacks on vulnerable iOS devices. Exposure does not, however, mean that every visitor’s data was stolen or their device compromised.

Third-party services provide valuable insights and accelerate development. Using them securely requires shared practices and clearly defined responsibilities. Procuring a service from a partner does not remove an organisation’s responsibility for the security of its own service and its customers.

According to Optimizely, the attacker exploited a vulnerability in the account creation process of its Experimentation service to inject malicious code into customer projects. There was no need to breach the organisations’ online services directly: the route ran through a trusted supplier.

How do we protect customers when an incident originates in a partner’s service?

Understand dependencies and their impact

Adding an analytics or testing tool to a website may be straightforward. Assessing its impact is not. When a supplier’s code runs in a customer’s browser, that supplier becomes part of the service’s security.

The Magecart and Ticketmaster cases demonstrated the role of third-party components in payment card data theft. The NotPetya attack highlighted an indirect risk: an organisation can become a victim through the technology it uses, even when it is not itself an intended target.

A good starting point is a shared understanding of dependencies: which third-party services are in use, what can they access, and who is responsible for the risks? A supplier register needs to be complemented by a joint impact assessment involving business and technical experts.

Threat modelling helps identify what could go wrong and how to limit the impact. A browser’s Content Security Policy (CSP) can restrict code execution, but its effectiveness depends on its configuration. Allowing code from a trusted supplier does not, on its own, protect against malicious code delivered through that supplier.

Security settings can affect how a website functions. Security specialists, developers and business teams using analytics therefore need to agree on security controls and their ownership together.

From compliance to security in practice

Requirements 6.4.3 and 11.6.1 of the PCI DSS standard address the authorisation of payment page scripts, integrity assurance and the detection of unauthorised changes. Their purpose is to protect customers’ payment data. Documentation needs to be supported by effective practices.

Every script on a payment page needs a justified purpose. The benefits of analytics must be weighed against customer security and the potential consequences of an incident. The solution may be to remove an unnecessary component or move payments to a payment service provider’s separate payment portal. An embedded payment form can also limit risk, but does not, on its own, remove security responsibilities for the surrounding page.

Detection and prevention complement each other: monitoring changes does not automatically stop an attack. It is therefore important to agree on who responds to alerts and how customer exposure is brought to an end. An attack does not, in itself, demonstrate non-compliance. It does, however, provide a reason to assess how security controls work in practice and where they can be improved.

Build security in together from the start

Collaboration between business teams, developers and security specialists helps align objectives. When security is integrated into procurement and design from the outset, solutions can be assessed against both user and business needs.

A slightly slower implementation or more limited analytics can be the right choice. This is not about giving up on development, but about safeguarding customer trust and business continuity.

Start with these three questions

  1. 1. Do we know which suppliers’ code runs in our online service and what it can do?
  2. 2. Have our business and security teams jointly assessed which components are genuinely necessary – particularly on payment pages?
  3. 3. Can we detect unauthorised changes, and are the responsibilities and procedures for responding to them clear?

You do not need to have every answer ready. Working through these questions together is a good first step. A clear understanding of dependencies, responsibilities and practices provides a foundation for developing cyber security systematically, as part of a trusted service and a seamless everyday customer experience.

Cybersecurity

Antti Laatikainen

Principal Cybersecurity Advisor

Antti Laatikainen is Gofore's Principal Cybersecurity Advisor. He has influenced e.g. a medical company's demanding operating environment as well as meaningful societal networks. Some 10 years ago Laatikainen moved onto consulting, where he has built compliant management models, strengthened cyber secure operating cultures and audited assessed security in several industries.

Back to top