0%
Security Engineering for Software Teams
Thinking About Security
Cryptography You Will Actually Use
Identity
Vulnerability Classes
Infrastructure and Cloud
Detection and Response
Compliance and Privacy Obligations
Software teams increasingly work under security and privacy obligations set by law, contracts and industry standards. Compliance can feel like paperwork, but most frameworks describe sensible security practices, and customers often require evidence of them before they buy. This section explains the main kinds of obligations, how they relate to each other and how engineers can meet them without derailing delivery.
Why compliance matters to engineers
Compliance obligations shape how systems store data, who can access production, how changes are reviewed and how incidents are reported. Engineers implement most of these controls, and their systems produce most of the evidence. Understanding the obligations helps engineers build compliant systems from the start instead of retrofitting controls before an audit.
Laws, contracts and standards
Obligations come from three main sources. Laws and regulations, such as data protection laws, apply because of where an organisation operates or whose data it handles. Contracts with customers and partners add commitments such as security requirements and notification times. Voluntary standards and certifications, such as security attestation reports and information security management certifications, demonstrate practices to customers.
Security attestation reports
Many business customers ask software providers for an independent attestation report describing security controls and whether they operated effectively over a period. The provider defines controls covering areas such as access, change management, monitoring and incident response, and an independent auditor tests them. A report covering a period of operation is more convincing than a point in time description.
Information security management certification
An information security management certification shows that an organisation runs a management system for security: assessing risks, selecting controls, monitoring them and improving over time. It is widely recognised internationally. The emphasis on risk assessment and continuous improvement makes it more about how security is managed than about specific technical settings.
Payment card standards
Organisations that store, process or transmit payment card data must meet industry card security standards. The requirements are detailed and technical, covering network segmentation, encryption, access control, logging and testing. Reducing the systems that touch card data, for example by using a payment provider hosted fields, dramatically reduces the scope and cost of compliance.
Health data rules
Organisations handling health information for healthcare providers or insurers may be subject to specific health data rules covering privacy, security and breach notification. These often require formal agreements with service providers, risk analyses and safeguards such as access controls and audit logs. Engineers building health related features should check early whether these rules apply.
Data protection laws
Data protection laws regulate how personal data is collected, used, stored, shared and deleted. Many share common principles: a lawful reason for processing, transparency, data minimisation, purpose limitation, accuracy, storage limitation, security and accountability. They give individuals rights such as access and deletion and impose breach notification duties and penalties for violations.
Overlap between frameworks
Frameworks overlap heavily. Access reviews, encryption, logging, change management, vulnerability management and incident response appear in almost all of them. Mapping controls to multiple frameworks lets one well designed control satisfy many requirements, avoiding separate programmes for each standard and reducing duplicate evidence requests.
Compliance is not security
Passing an audit does not mean a system is secure. Audits test whether defined controls exist and operate, often by sampling. Attackers do not care about audit scope. Organisations should use compliance as a floor and a forcing function for good practices, while continuing to address risks that frameworks do not cover or cover only loosely.
Security supports compliance
The reverse is also true: a strong security programme makes compliance easier. Organisations that already manage access tightly, automate infrastructure, review changes and monitor systems can produce evidence quickly. Compliance then becomes documentation of existing good practice rather than a scramble to create controls.
Scope
Every obligation has a scope: which systems, data, locations and people it covers. Narrowing scope through architecture, such as isolating card processing or health data into separate environments, reduces cost and complexity. Engineers influence scope through design decisions, which is one reason compliance should be considered early in projects.
Shared responsibility with cloud providers
Cloud providers hold their own certifications and reports covering their infrastructure. Customers inherit some controls, such as physical security, but remain responsible for how they configure and use cloud services. Understanding which controls are inherited and which are the customer responsibility avoids gaps and duplicate work.
The engineer role
Engineers design systems with appropriate controls, implement access and change management, configure logging and encryption, respond to evidence requests and fix control failures. They also flag when a new feature might create new obligations, such as collecting a new type of personal data or expanding into a new region.
Questions engineers should ask
When starting a feature, engineers can ask: what personal or sensitive data does this touch, where will it be stored and processed, who will have access, how long is it kept, which customers or regions are affected and does it change any controls that are in audit scope? These questions catch most compliance impacts early.
Common misconceptions
Common misconceptions include believing that using a compliant cloud provider makes an application compliant, that encryption alone satisfies data protection laws, that compliance is the security or legal team problem alone and that certification is a one time project. Each leads to gaps that surface during audits or incidents.
Who decides that an obligation applies
Engineers rarely decide alone whether a law or standard applies. Legal, privacy and compliance specialists interpret obligations, while engineers describe what systems actually do with data. The most useful engineering input is precise: which fields are collected, from which users, in which regions, for what purpose and shared with which services. Precise descriptions lead to accurate legal conclusions, while vague ones lead to either overcautious restrictions or missed obligations.
Customer contracts in detail
Enterprise contracts often contain security schedules listing required controls, such as encryption standards, background checks, penetration testing frequency, data location and incident notification times. Sales teams sometimes agree to terms that engineering cannot meet without notice. A standard security schedule, reviewed by engineering and offered by default, prevents surprises and keeps commitments consistent across customers, which makes them far easier to evidence later.
Regulated industries
Customers in finance, healthcare, government and critical infrastructure often pass their own regulatory obligations down to suppliers. A bank customer may require specific audit rights, resilience testing or outsourcing notifications. Selling into these sectors can therefore add obligations that no law imposes on the supplier directly. Product leaders should understand these requirements before targeting a regulated sector, because they affect architecture, support and cost.
The cost of getting it wrong
Failing obligations has several costs: regulatory fines, contractual penalties, lost deals when evidence is missing, mandatory remediation programmes and reputational damage after publicised failures. The largest cost is often lost trust from customers who chose the provider partly because of its security claims. Investing in controls early is usually cheaper than responding to a failure, especially when the failure becomes public.
Proportionality
Most frameworks expect controls proportionate to risk and size. A small startup does not need the same programme as a global bank, but it does need the essentials: strong access control, encryption, logging, reviewed changes, incident response and honest data handling. Starting with these essentials and growing the programme as the business and its obligations grow avoids both neglect and paralysing bureaucracy.
Compliance roadmaps
Early stage companies often face their first certification request from a large prospect. A roadmap that sequences readiness work, picks an audit period, chooses tools and assigns owners turns a vague demand into a plan. Engineering can usually meet most requirements with a few months of focused work if the systems already follow good practice, which is the strongest argument for building well from the beginning.
Laws follow the data you handle and where your users are, contracts add customer commitments and standards demonstrate practices. Map one set of controls to many frameworks, narrow scope through architecture, understand inherited cloud controls, treat compliance as a floor rather than proof of security and ask early whether a feature creates new obligations.