Our information security policy
Doroobi is made and run by ITD Innovation Information Technology in Dubai. This is the public summary of ITD’s information security policy, ITD-ISMS-001. It sets the rules that ITD’s team, contractors and products follow.
Version 1.3 · 18 November 2026 · reviewed every year, and after any major incident
This page describes the policy: the rules we hold ourselves to. What is already in place for Doroobi is on the Security page, which says only what is true today.
Scope and responsibility
- Everyone and everything. The policy covers every person who works for ITD (staff, managers, contractors, freelancers and interns) and every ITD product, service, server, account and set of data, our own and our clients’.
- Clear owners. The Managing Director approves the policy and decides on remaining risks. A Security Lead owns it, approves access and leads the response to incidents. Every person must follow it and report anything suspicious at once.
- Built on a standard. The policy follows the structure of ISO/IEC 27001, so that ITD can work towards certification. ITD is not certified today.
Law and privacy
- The laws we work under. The UAE Personal Data Protection Law (Federal Decree-Law No. 45 of 2021) and Cybercrime Law (Federal Decree-Law No. 34 of 2021); where they apply, the DIFC and ADGM data protection rules, Saudi Arabia’s Personal Data Protection Law, the EU GDPR and the app stores’ rules. For each significant project a licensed lawyer confirms what applies.
- Privacy by design. We collect only the data a feature needs, say in a privacy notice what we collect, why, where it is kept and for how long, ask for consent where the law requires it (above all for location, marketing and cookies), and delete or anonymise data when its time is up.
- Location is high-risk data. Live location and tracking of employees need clear consent, a limited purpose and a limited time they are kept.
- Your rights. You can see, correct and delete your data, and our apps let you delete your account from inside the app.
- Breaches. A personal data breach is reported inside ITD at once and assessed for telling the authorities, clients and the people affected within the deadlines of the law and our contracts.
How we classify data
- Public. Approved website and store content.
- Internal. Our procedures and plans: ITD staff only.
- Confidential. Contracts, source code, configurations and client business data: on a need-to-know basis, encrypted when stored and sent, under a confidentiality agreement.
- Restricted. Passwords, keys, backups and personal data such as live location: named people only, two-step sign-in, encrypted, every access logged, and never in chat, email or AI tools.
Access and accounts
- Least privilege. Each person gets only the access their role needs, for only as long as needed.
- Two-step sign-in everywhere it matters. Email, cloud and domain accounts, code repositories, app stores, social and business accounts, payment and banking. Authenticator apps or security keys rather than text messages.
- One person, one account. No shared accounts or passwords. Company accounts are registered to ITD, never to a person’s private account.
- Passwords. A company password manager; at least 14 characters or a passphrase, different for every system, never sent by WhatsApp, email or text message, changed at once if it may have been exposed.
- Joining and leaving. Access is approved and logged. When someone leaves, all their access ends on their last working day. Access to critical systems is reviewed every quarter.
How we build software
- A secure development process. Security and privacy requirements in every specification, a threat model for new products, a second person reviewing every change, automated tests with code, dependency and secret scanning, and staging before production. We work to the OWASP Top 10, OWASP API Security Top 10, ASVS and MASVS.
- The server decides. Every request is checked on the server: who is asking, their role and their company. Hiding a button is never the protection. Tests call every route as the wrong person and expect a refusal.
- Companies are kept apart. In products shared by many companies, every query is limited to one company, and automated tests prove that one company cannot read, change or export another’s data, on every release.
- No secrets in apps or websites. Keys stay on servers. Anything that ships in an app or a web page is treated as public.
- Sign-in done properly. Passwords kept only as slow, salted hashes; two-step sign-in for administrators; limits on repeated failures; sessions that end and that are ended on the server at sign-out and when a password changes; passwords and tokens never written to logs. Social logins use the authorisation code flow with PKCE in the phone’s own browser.
- Mobile apps. Sign-in tokens in the phone’s secure storage (Keychain or Keystore), only the permissions a feature needs, and location while the app is in use unless more is truly needed.
- AI-directed engineering. AI tools give speed; our engineers give the judgement. Before any customer uses a system, its builder answers in writing: where does the data live, does the server check who may do what, and where are the keys. Every product goes through a security hardening audit before launch and every year, and is scanned with OWASP ZAP before launch, never against live systems.
Servers, networks and email
- Encrypted on the way and at rest. TLS 1.2 or newer everywhere; laptops, servers, databases and backups encrypted.
- Hardened servers. Firewalls that deny by default, no database or admin ports open to the internet, publishing through a reverse proxy or Cloudflare Tunnel, automatic security updates and monitoring.
- Behind an edge. Public sites sit behind Cloudflare, with rate limits on sign-in and forms.
- Email that cannot be faked. SPF, DKIM and DMARC on ITD’s domains.
Backups and recovery
- 3-2-1. Three copies, on two kinds of storage, one of them away from the server, and at least one that cannot be changed or deleted. Backups are encrypted and their keys are kept apart.
- Targets set by the business. For each system, how long it may be down (RTO) and how much data it may lose (RPO); production databases can be restored to a point in time.
- Restores that are tried. At least every quarter, a backup is restored into a clean place, checked and timed against those targets.
Vulnerabilities and incidents
- Fix deadlines. Critical: 48 hours, with a stopgap the same day. High: 7 days. Medium: 30 days. Low: 90 days or the next release.
- Testing. Monthly scans of public systems; an external penetration test every year and before major launches.
- When something goes wrong. Four levels of severity. We contain first, keep the evidence, remove the cause, restore from known-good backups, change every exposed key, tell clients, users and authorities as the law and our contracts require, and review what happened within five working days.
- Tell us. Write to [email protected]. The address is also in our security.txt.
Suppliers and AI
- Suppliers checked first. Before we use a supplier we check its security, where it keeps data, its certifications and its data processing agreement. Integrations use official APIs, with keys limited to what they need and changed when people change.
- AI tools used by our team. Never with passwords, keys, personal data or other restricted data. What AI produces is reviewed like any other work, and people stay accountable for it.
- AI inside our products. No more power than the person using it, protection against prompt injection, personal data sent to an AI provider only when needed and disclosed, every action logged, and a person’s approval for anything with real effect, such as sending a campaign.
People and devices
- Training. Security training when people join and every year; confidentiality in every employment and contractor agreement.
- Devices. Encrypted disks, automatic updates, antivirus and firewall on, screens that lock; a lost or stolen device is reported at once and wiped remotely.
- Premises. Office access for staff and escorted visitors; servers, network equipment and backup drives locked away.
This summary is not a certification, an audit report or legal advice. Questions about it: [email protected].
See also the Security page and the privacy policy.