An example scenario: a company built a product with 100,000 users. During due diligence before a funding round, it discovered a huge privacy exposure in its data architecture, and it lost the round. Privacy by design is not just “the right thing to do”. It is a matter of business survival.

What is privacy by design?

Privacy by Design (PbD) is an approach to product development in which privacy protection is built into the system from day one, not added afterwards. The idea was developed in the 1990s by Dr. Ann Cavoukian of Canada. It became a legal duty in 2018, when the European GDPR came into force (Article 25).

In Israel, the Protection of Privacy Regulations (Data Security), 5777-2017 go in the same direction. They require organizations to adopt security measures suited to the sensitivity of the information and to the risks. In other words, to think of privacy as part of the design, not as a checkbox.

The seven principles of privacy by design

Cavoukian set out seven guiding principles. This is how they look in practice in a technology company:

  1. Prevention, not correction: identify privacy risks before development, not with a check after the code is in production. Threat modeling at the design stage.
  2. Privacy as the default: the product shares as little information as possible and requires active consent to share more. The default is “don’t share”, not “share everything”.
  3. Privacy built into the architecture: privacy protection is part of the technical design, not an external plugin. Encryption, network separation and minimum access are planned in advance.
  4. Not zero-sum: privacy and security do not conflict with functionality. You can build a product that does both. “It will hurt the user experience” is usually an excuse, not a real limitation.
  5. End-to-end security: the data is protected from the moment it is collected until it is deleted. Not only in transit, but also in storage, in backups, and in every internal interface.
  6. Transparency and openness: users know what is collected, why, and how. Not a 40-page privacy policy that nobody reads, but real transparency.
  7. Respect for user privacy: sensitive information gets more protection, consents can be withdrawn, and users can access and delete their data.

A question about a website, privacy, terms of use or AI in your business?I will get back to you within one business day.

Calls and replies in English or Hebrew

What happens after you write to me

  1. You write what happened and with whom. No documents.
  2. I get back to you within one business day for a short call of a few minutes.
  3. If it is a matter I handle: after the call you will receive, in writing, what will be done and what the next step is. Work starts only after you approve.

At the end of the call you will know one of three things: it is a matter I handle, and what the next step is; you can manage without a lawyer, and which guide fits; or it is outside my field, and where to turn.

Practical implementation in software development

Four tools that every development team should adopt:

Privacy Impact Assessment: when and why

A DPIA (Data Protection Impact Assessment) is a process for assessing how a new project will affect privacy. Under the GDPR it is required before processing that is likely to result in a high risk. Examples are large-scale processing of special categories of data, such as biometric or medical data, or systematic and extensive profiling (Article 35). In Israel it is not required by law, but it has become an important legal protection tool.

A DPIA includes: a description of the data processing, an assessment of necessity and lawfulness, identification of the risks to data subjects, and mitigation measures. A company that carried out a DPIA and documented it is in a much better position before a regulator and before a court, if a breach occurs.

What happens to companies that don’t implement it?

Three main exposures:

Has something similar happened to you?You can write to me what happened and with whom, without documents.

Calls and replies in English or Hebrew

A checklist for a company that wants to start

Five practical steps for a company that is starting to implement it:

  1. Appoint a DPO (Data Protection Officer). Since Amendment No. 13 this is a duty for public bodies, for data brokers, and for those whose main activity includes systematic monitoring or processing sensitive information on a significant scale (section 17B1; see who must appoint an officer after Amendment No. 13). In all other cases, at least an internal person in charge. Someone has to own the subject.
  2. Create a Data Inventory: a list of all the personal information you hold, where it came from, and where it is kept.
  3. Carry out a DPIA for new projects: before the sprint, not after it.
  4. Write a Privacy Policy that people can read: short, clear and accurate.
  5. Train the development team. A developer who knows what PbD is will make better decisions without waiting for a lawyer.

Frequently asked questions

Is a privacy impact assessment (DPIA) mandatory in Israel?

In Israel it is not required by law, but it has become an important legal protection tool. Under the GDPR it is required before processing that is likely to result in a high risk. Examples are large-scale processing of special categories of data, such as biometric or medical data, or systematic and extensive profiling (Article 35).

We have a good privacy policy. Is that privacy by design?

No. The policy is a document; PbD is architecture. You can have an excellent policy alongside a completely broken architecture.

Who must appoint a privacy protection officer (DPO)?

Since Amendment No. 13 this is a duty for public bodies, for data brokers, and for those whose main activity includes systematic monitoring or processing sensitive information on a significant scale (section 17B1 of the Protection of Privacy Law). In all other cases, it is worth appointing at least an internal person in charge.

Sources

The links lead to the original legislation (in Hebrew). The binding text is the one published in Sefer HaHukim (the official gazette of statutes).

Checklist

  • Practical tip: in a technology company, prepare a Data Flow Diagram (a map of how the data flows) before any privacy review. Without knowing where the data goes, any privacy review is superficial. If no DFD can be presented, that is a finding in itself.
  • Relevant legislation: Article 25 of the GDPR (Data protection by design and by default); Protection of Privacy Regulations (Data Security), 5777-2017; Protection of Privacy Law, 5741-1981.
  • Common mistake: thinking that a good privacy policy equals privacy by design. The policy is a document; PbD is architecture. A company can have an excellent policy and a completely broken architecture.
  • Tactical point: in a review before a product launch, carry out a documented DPIA, even if it is not mandatory. Documentation showing that a company recognized the risks and did everything to reduce them can serve as evidence of reasonable conduct if a claim is filed.
Legal note: The information in this guide is general and is not legal advice or a substitute for advice on your own case. Every case is assessed on its own facts.