Skip to content

How We Build · 6 min read
Why we plan security before we build anything

Nobody builds an office and then asks where the doors and locks should go. Yet software is often built that way: the features come first, and security is a line near the end of the schedule, tested in the last week if time allows. For a system that holds payroll, health records or the names of beneficiaries, that order is the wrong way round.

Who carries the risk

The United States Cybersecurity and Infrastructure Security Agency (CISA) argues that the burden of security sits in the wrong place. It says the burden falls disproportionately on consumers and small organizations, and that every technology provider must take ownership of making its products secure.1

Its answer is a principle called secure by design. A product built this way treats the security of its customers as a core business requirement and not just a technical feature. The agency says the work belongs in the design phase, where it can greatly reduce the number of flaws before a product reaches the public, and that protections such as two-step sign-in and activity logging should come in the box at no extra cost.1

The guidance, first published in April 2023 and updated with 17 partner agencies that October, sets three principles for software makers. The first is to take ownership of customer security outcomes.2 We agree with it. A ministry or a cooperative hires a developer because it does not have security engineers of its own. Handing over a system that is safe only if the client configures it expertly is not a finished job.

Thinking like the person who wants to break in

The first security task on a project needs no code. It is called threat modeling, and it means working out what could go wrong before deciding how to build. The Open Worldwide Application Security Project (OWASP), a nonprofit that publishes free security guidance, reduces it to four questions.3

The questions are: what are we working on, what can go wrong, what are we going to do about it, and did we do a good job.3 For a scholarship system, the second question produces a concrete list. An applicant might try to view another applicant's documents. A staff member might change a score. Someone might guess a weak password or send a fake link to an administrator.

OWASP advises starting this early, at the planning stage, and returning to it throughout the project.3 Each item on the list becomes a design decision while changing the design is still cheap.

  • What are we working on?
  • What can go wrong?
  • What are we going to do about it?
  • Did we do a good job?

Giving each person only the access the job needs

The second habit is least privilege. The United States National Institute of Standards and Technology (NIST) defines it as designing a system so that each user or program is granted the minimum resources and authorizations needed to perform its function.4

In an office this is familiar. The cashier holds the cash box key, not the key to the records room. In software it means an encoder can add records but not delete them, a division chief sees her own division, and nobody uses a shared administrator account for daily work. When an account is stolen or misused, the damage stops at the edge of what that account could do.

This has to be decided early because it shapes the whole system. Roles and permissions are built into every screen and every stored record. Adding them to a finished system that assumed everyone could see everything means reopening most of it.

The ten most common ways systems fail

OWASP also publishes the Top 10, a standard awareness document that represents a broad consensus on the most critical security risks to web applications. The current edition is dated 2025.5

At the top of the list is broken access control, which is the failure of least privilege: a user reaching data or actions that should be closed to them. Second is security misconfiguration, such as a system left on its default settings. Third is software supply chain failures, which covers weaknesses in the outside components a system is built from. Insecure design has its own place on the list, at number six.5

Our approach is to use the list as a checklist during design and again during testing. It is not a guarantee. It is a record of the mistakes other teams made most often, and there is no good reason to repeat them.

Why late security costs more

A study prepared for NIST in 2002 examined the cost of software defects in general, not only security flaws. It concluded that early detection can greatly reduce costs, and that the longer a defect stays in a program, the more costly it becomes to fix. It estimated the yearly cost of inadequate testing to the United States economy at 22.2 to 59.5 billion dollars, with over half borne by the users of the software and not its makers.6

Security flaws follow the same pattern, with an added cost: a flaw found after launch may already have exposed people's records. NIST's framework for secure development makes the point that few development methods address security in detail, so security practices usually have to be added to whichever method a team follows. It says that following such practices should help reduce the number of weaknesses in released software and address their root causes.7

Philippine law puts the duty on the office, not only the supplier. The Data Privacy Act of 2012 requires anyone who controls personal information to implement reasonable and appropriate organizational, physical and technical measures to protect it, and makes the head of each government agency responsible for complying with its security requirements.8

What an office can ask before signing

You do not need to be technical to tell whether a supplier takes this seriously. Ask to see the list of what could go wrong with your system and what was decided about each item. Ask for the table of roles: who can view, add, change, approve and delete. Ask whether two-step sign-in and activity logs are included or sold as extras.

Ask one more thing: what happens when a flaw is found after launch, who fixes it, how fast, and who pays. Our approach is to settle these questions in the first weeks of a project and write the answers down, so that the office holding the data knows what it is getting before anyone's records are in the system.

Sources

  1. Secure by Design. Cybersecurity and Infrastructure Security Agency (CISA).
  2. Shifting the Balance of Cybersecurity Risk: Principles and Approaches for Secure by Design Software. Cybersecurity and Infrastructure Security Agency (CISA) and partner agencies, 2023-10-25.
  3. Threat Modeling. OWASP Foundation.
  4. least privilege - Glossary. NIST Computer Security Resource Center.
  5. OWASP Top 10:2025. OWASP Foundation, 2025.
  6. The Economic Impacts of Inadequate Infrastructure for Software Testing, Final Report. RTI, prepared for the National Institute of Standards and Technology, May 2002.
  7. Secure Software Development Framework (SSDF) Version 1.1: Recommendations for Mitigating the Risk of Software Vulnerabilities (NIST SP 800-218). National Institute of Standards and Technology, February 2022.
  8. Republic Act No. 10173, Data Privacy Act of 2012. The Lawphil Project, Arellano Law Foundation, 2012-08-15.
#HowWeBuild#BARMMTech#MoroTech
Share