- Marriott identified about 383 million guest records as the upper limit involved in the Starwood incident; MGM estimated a US$100 million hit to Adjusted Property EBITDAR after its 2023 attack
- The main exposure points are the PMS, payment systems, guest Wi-Fi, booking channels and OTA accounts, staff identities and supplier remote access
- Network segmentation matters, but it works alongside phishing-resistant MFA, vulnerability management, payment security, supplier control and incident response
- Controller and processor roles depend on each processing activity; OTAs generally act as independent controllers for their own platforms
- Under UK GDPR, a breach must be reported to the ICO within 72 hours unless it is unlikely to result in a risk to individuals
Why hotels attract attackers
Marriott disclosed the Starwood guest reservation database incident in 2018. After its investigation, the company identified approximately 383 million guest records as the upper limit of records involved, and stressed that this did not mean 383 million unique guests, because many guests appeared in the database more than once (Marriott update, January 2019). In September 2023, a cyberattack on MGM Resorts disrupted systems across its properties. In a filing with the US Securities and Exchange Commission, MGM estimated a negative impact of approximately US$100 million on Adjusted Property EBITDAR for its Las Vegas Strip and regional operations, plus less than US$10 million in one-time expenses for the quarter (MGM Form 8-K, October 2023).
Several factors tend to coincide in hotels. They handle different categories of personal data (identity details, contact data, stay history, payment information). They run many connected systems from different suppliers. Front desk teams change often and work under time pressure. Operations run around the clock, so any outage is felt by guests immediately. None of this makes a breach inevitable, but together these factors explain why criminal groups and other threat actors keep targeting the sector.
The main attack surfaces in a hotel
Hotel environments often combine guest networks, operational systems (PMS, POS, door locks, CCTV, building management), supplier devices and corporate applications such as email and Microsoft 365. Where segmentation is insufficient, a compromise in one area can reach the others and widen the exposure. The practical question is not whether VLANs exist, but which communication paths are actually allowed between zones and whether anyone has tested them.
Property management systems
The PMS holds guest profiles, reservations, room allocation, folios and, depending on the setup, payment tokens. What an attacker can reach after compromising a PMS account depends on that account's privileges and on the architecture. A cloud PMS with role-based access and tokenised payment data exposes far less than an on-premises installation where a shared administrator account can export the entire guest database. Outdated installations do exist and deserve attention, but many hotels run current or vendor-hosted versions, so the patch status has to be checked rather than assumed.
The core controls are role-based access, so that front desk users see only what they need; individual accounts with strong authentication, with MFA as a minimum for administrators and remote access; protection of integrations such as API keys, interface accounts and service accounts; and monitoring of exports, privilege changes and unusual logins.
POS and payment systems
POS malware campaigns have hit hotel restaurants and bars in the past, capturing card data on compromised terminals or servers. How exposed a hotel is today depends on its payment architecture: certified payment terminals, point-to-point encryption (P2PE), tokenisation, integrated payment gateways and the actual flow of card data all change what an attacker could capture. Following PCI DSS scoping, start by identifying every component that stores, processes or transmits cardholder data, or that can affect its security, and keep that environment as small and isolated as possible. Our guide to PCI DSS for hotels covers scoping in detail.
Guest Wi-Fi
Guest Wi-Fi should give guests internet access and nothing else. Risk arises when guest traffic can reach internal systems, management interfaces or other guests' devices. VLANs on their own are not enough. What matters is routing controls, firewall policies that deny traffic from guest networks to internal zones by default, client isolation between guests, and periodic testing that confirms the separation works as designed. Our article on guest Wi-Fi network segmentation in hotels sets out a step-by-step approach.
Booking engine, OTAs and email
Hotels receive bookings through their own website and through online travel agencies (OTAs) such as Booking.com and Expedia, usually via a channel manager. In recent years attackers have repeatedly targeted the accounts rather than the integrations. A message posing as an OTA or a guest asks front desk staff to log in or open a file, and the stolen credentials are then used to access the OTA extranet or a mailbox and send guests fake payment requests. Protect these accounts with MFA (phishing-resistant methods where the platform supports them), individual user accounts instead of shared logins, a secure account recovery procedure and a fast way to revoke access when staff leave.
Third-party remote access
Hotels depend on suppliers who often need remote access: PMS integrators, POS providers, door-lock vendors, Wi-Fi contractors and outsourced IT. Every remote connection is a possible entry point. Give each supplier named accounts rather than shared ones, require MFA, grant the least privilege needed, make access time-limited where feasible (enabled on request for a defined window), log sessions and revoke access when a contract ends or an engineer leaves the supplier.
A compact hotel security assessment
The table below summarises the areas to review in a hotel, the main concern in each and examples of evidence that show whether the controls actually work.
| Area | Main concern | Example evidence |
|---|---|---|
| PMS | Excessive privileges, shared accounts, unprotected integrations | User and role list, MFA settings, integration inventory, export logs |
| POS and payments | Card data captured or stored outside protected components | Card data flow diagram, P2PE or tokenisation confirmation, PCI DSS scope document |
| Guest Wi-Fi | Guest traffic reaching internal systems or other guests | Firewall rules between zones, client isolation settings, segmentation test results |
| Booking engine and OTAs | Account takeover leading to guest fraud | OTA user list, MFA status, recovery procedure, alert settings |
| Identity and Microsoft 365 | Credential theft through phishing, weak protection of admin accounts | Conditional access policies, MFA coverage report, admin role review |
| Physical access and IoT (locks, CCTV, smart rooms) | Default credentials, flat networks, unsupported firmware | Device inventory, firmware versions, network placement, supplier access list |
| Backup and recovery | Backups encrypted by ransomware or never restored | Offline or immutable copies, restore test records, recovery time targets |
| Suppliers | Uncontrolled remote access and unclear responsibilities | Supplier register, security clauses in contracts, remote access logs |
UK GDPR obligations to get right
Controller and processor roles are not fixed by the type of supplier. They depend on who determines the purposes and essential means of each processing activity. A PMS provider usually acts as a processor for the guest records it handles on the hotel's instructions, but it can be a controller for its own purposes, such as billing its customers. OTAs such as Booking.com and Expedia generally act as independent controllers for the processing on their own platforms, where they have their own relationship with the guest and their own privacy terms. Channel managers, booking engines and restaurant reservation tools need the same assessment, activity by activity. Contracts and data-sharing arrangements, whether a processing agreement under Article 28 or a controller-to-controller arrangement, must reflect the real roles. Our GDPR compliance guide explains the distinction.
Retention should be defined per data category and purpose, not as one period for all guest data. Document each period and the reason for it, review it regularly and delete or anonymise the data when the period ends.
| Record type | Retention considerations |
|---|---|
| Registration data | Guest registration requirements set which details are needed and for how long; keep only those fields |
| Tax invoices and accounting records | Statutory tax and accounting retention periods apply |
| Payment records | Sensitive authentication data must not be stored after authorisation; any stored card numbers must be protected under PCI DSS; transaction records follow accounting needs |
| Guest communications | Keep while needed to handle the stay, complaints or claims, then delete |
| Loyalty programme data | Keep while the membership is active, with a defined period after inactivity |
| Passport or ID images | Only where necessity and a legal basis are established; short, justified retention and a documented deletion procedure |
Passport and ID copies need a careful assessment. Collecting and keeping images of identity documents must be justified by purpose, necessity and legal basis. Needing certain identity details, for example the guest records required under the Immigration (Hotel Records) Order 1972, does not by itself justify keeping a full image of the document. If a hotel concludes that scanning is necessary, it should restrict access to the images, set a justified retention period and follow a documented deletion procedure, subject to any legal obligation or legitimate need to keep them longer. A scan saved in a shared front desk folder meets none of these conditions.
Under UK GDPR Article 33, a personal data breach must be reported to the ICO without undue delay and, where feasible, within 72 hours of becoming aware of it, unless the breach is unlikely to result in a risk to individuals' rights and freedoms. Where the breach is likely to result in a high risk, Article 34 also requires informing the affected guests without undue delay. Hotels without a tested notification procedure face a greater risk of delay, because the assessment has to be made while the investigation is still running.
Sanctions depend on the infringements the regulator establishes, their gravity and the procedure followed. An incident does not automatically lead to a fine, but a poorly documented response makes it harder to show that the hotel acted appropriately. See our guide to incident notification duties for the wider picture.
Initial security priorities for hotels
The six controls below are a practical starting point for hotels of any size. They are initial priorities, not a complete baseline. Network segmentation is important, but it is not always the single most effective control: identity protection, vulnerability management, payment security, supplier control and incident response often matter just as much, so weigh them together.
Operational continuity deserves its own scenario. If the PMS is unavailable, how does the front desk see today's arrivals, allocate rooms and take payment? Typical answers include offline procedures and forms, a recent export of arrivals and in-house guests stored securely, payment terminals that can work independently of the PMS, a door-lock system that can issue keys locally where it supports this, and a PMS restore that has actually been tested within an agreed time.
Beyond these priorities, a complete programme also covers backup and recovery with protected copies and tested restores, privileged access management, supplier and API security, an up-to-date asset inventory, endpoint and server protection, a defined PCI DSS scope, security monitoring, and joiner, mover and leaver processes so that access ends when staff or suppliers leave.
This article is an executive overview. For more depth, see our hotel cybersecurity risk report 2026, the guide to guest Wi-Fi segmentation, PCI DSS for hotels, the GDPR compliance guide and our incident response plan guide.
Sources: Marriott, update on the Starwood database security incident (4 January 2019); MGM Resorts International, Form 8-K (5 October 2023); ICO, reporting a personal data breach; PCI Security Standards Council.