This week I want to address an important issue for the hospitality industry, albeit one that technologists have not historically had on their radar. They can potentially play a significant role in addressing the issue, which is human trafficking, and more specifically sex trafficking.
Only a few of the capabilities described here have been implemented, mostly as custom modifications that cannot be purchased commercially. To be clear, the systems deployed in hotels can often capture the information that is needed, but they need enhancements (mostly minor) that have not been written. Cooperation across systems is needed, but hotel systems are not integrated in ways that could help address the problem.
I will come back to the technology piece in a minute, but I will start by framing the issue with some background on trafficking and hotels. According to a Human Trafficking Institute study, 80% of new federal criminal sex trafficking cases filed in the U.S. in 2021, where the location of the sex act was known, were in hotels. The same report showed that hotels were by far the most common defendants in new civil sex trafficking lawsuits, with more than twice as many actions as the runner-up category. Similar numbers emerge from the National Human Trafficking Hotline, which according to a Polaris study had 3,596 reports of human trafficking involving a hotel or motel; in 81% of these cases, sex acts were performed in a hotel under an escort service business model. Trafficking is not limited to one category of hotel; it happens in every category from bare-bones budget to top-end luxury.
Much has been written about this problem, and it is not my purpose as a technology writer to cover it in depth. Suffice it to say that the research shows that most victims of sex trafficking were lured or tricked into it by criminal traffickers, or forced by someone on whom they depend. Traffickers use physical and psychological threats to convince the victims that they are trapped with no way out. Victims are often afraid to speak up due to fear, isolation, guilt, shame, or manipulation by traffickers. Research from OnWatch indicates that 40% of traffickers are related to their victims, often uncles or more distant relatives but sometimes even parents. Many trafficking victims lack the support network from family that may be needed.
Alert hotel staff and guests who see something can call a national hotline operated by the nonprofit group Polaris. The hotline can help assess the situation and take appropriate action. A national campaign by the AHLA Foundation has prioritized actions around awareness, training, and posters for hotels, as well as survivor support. Posters are now appearing in many hotels encouraging guests and staff to make the hotline call.
A free 30-minute training program was developed collaboratively by Marriott International, ECPAT USA, and the AHLA Foundation. This is recommended or required training for many hotel companies. More than 1,000,000 hotel staff at Marriott alone have now completed it, and many more at other hotel companies. If you have not studied sex trafficking before, taking the training will change your perspective significantly.
The Hotel Industry and Trafficking
Most hotel companies have been tight-lipped about efforts to combat trafficking, a factor that significantly impeded my research. There is an implied belief that if you talk about how to fix a problem, it’s an admission that there’s a problem. Of course, denying the problem makes it unlikely hotels will be able to fix it.
Trafficking in hotels is indeed a sensitive topic but is finally starting to get some visibility. Public shaming and lawsuits, as well as campaigns by various organizations, have forced the issue into the open. Marriott (the only hotel company I found that was willing to be interviewed about trafficking, even off-the-record) launched an awareness training program around 2016 and has continually updated and expanded it, contributing material that they are sharing with other hotel groups as well. I applaud them for taking the leadership. Signage promoting the Polaris hotline has also appeared in public spaces of many hotels.
But hotels’ lack of willingness to talk about the issue has meant that many things that could be done, have not. Tech vendors are not going to spend money to solve problems that their customers deny even exist. Tools that could enable hotels to fight trafficking are still on the idea shelf or in their infancy. Most of the efforts to date have focused on training staff and managers. This is critical, as staff are the ones that may detect suspicious behavior and that will be called upon to take appropriate action. But it is not enough.
As an industry, hotels do very little to harness the technologies they already have to help identify trafficking. Many known behaviors of sex traffickers can be detected and analyzed electronically, but with few exceptions this has not been done. It is, however, essential, because traffickers working in hotels know how to minimize the risk of being caught by front desk staff, housekeepers, and security personnel. To be sure, there are thorny privacy issues that need to be considered – but there are ways to address them.
The Role of Technology
Technologists can help, and a few activities are already underway. Marriott announced earlier this year a project with Internet Watch Foundation, Cisco, and its various Internet access providers to block access to known child sexual abuse websites from hotel networks (sites guests may visit to arrange liaisons). It is not perfect, since guests can use their mobile network or a virtual private network (VPN) to bypass the restrictions, but it is certainly helpful.
Training programs also help, but they need to be integrated into existing learning management systems to maximize the ability for hotels to measure compliance and to ensure consistent completion of the module during new-hire onboarding and recurring training. The Marriott/ECPAT program can be licensed to support this deployment model.
In-room entertainment systems can also be set up with a channel (or a segment on the hotel’s marketing channel) to highlight the same information as on printed posters about where/how to report suspicious activities. A few hotels have reportedly done this, another useful step.
A more interesting opportunity, which is only beginning to be developed, is to gather signals of potential trafficking from both hotel systems and human observation. This requires lots of data and analytics.
Traffickers (and their victims and clients) have usage and behavior patterns that, while not universal, are common enough to have been identified. The challenge, of course, is that these behaviors do not identify trafficking, but rather possible trafficking. Many everyday guests do the same things, and that does not make them traffickers. But a combination of multiple “signals” occurring more often than with typical guests may indicate that something is amiss and should be investigated.
This is a probability game that is ripe for analytics: hotels should be able to receive alerts when the likelihood of trafficking by a particular guest or in a particular room exceeds some threshold. The explanation may still be innocuous, but hotel staff can often investigate discreetly, or at least put colleagues or security systems on alert to watch a particular room or person. This can enable the hotel to monitor the guest without confronting them.
Privacy is a concern any time hotels collect data specific to individuals, and work needs to be done to determine how to assess the risk that a particular guest is a trafficker, without violating legal or ethical privacy guidelines. However, just as banks (and some hotels) collect and use data from multiple sources to assess and score credit risk, it should be possible to do the same for crime risk from guests while still protecting privacy.
Hotel systems can also collect evidence that can be used by law enforcement and prosecutors to both apprehend and prosecute traffickers. For this, IT policies on record retention are critical. While many records may be routinely purged after 30 days or some similar interval, longer retention of certain records may be requested by law enforcement particularly during active investigations. And with today’s low costs for cloud storage, even video recordings can be retained for years at minimal cost.
Requests from law enforcement may come on short notice to almost any hotel staff member (at the hotel or at a brand contact center), and immediate access to data may be needed. IT and operating policies need to ensure that every staff member knows what to do and who to contact when such a request comes in, for example to view security recordings or to search the property management system for a particular name or credit card. Many law enforcement investigations go cold because hotel staff are not trained on how to provide real-time access to their systems, or how to extend purge dates of archival records (assuming that is even possible). Clear policies should be in place as to what can be released to law enforcement on request vs. requiring a court order, and a designated set of individuals at a network operations or other contact center should be available 24x7, with the authority and capability to help front-line staff respond to such requests appropriately.
Digital Signals that May Indicate Sex Trafficking
In today’s hotels, digital events are captured by various hotel systems on a continuous basis, and some of them may indicate the possibility or likelihood of sex trafficking.
Starting with the booking, various factors can correlate with trafficking, including booking channel, requested room type, special requests, the size of the party, and the method of payment. In some countries where names and ages of accompanying guests are provided at booking, hotels may also be able to consider this information.
I am aware of one hotel company that flags for closer review any bookings that meet certain criteria. I was not able to get specific information as they declined an interview, but based on what I did learn, I suspect they use statistics collected by law enforcement to calibrate a booking-risk algorithm. This may even include watch lists for certain names or credit cards. For most flagged bookings, the likely action would be to disable mobile check-in (if offered) and to alert the front desk clerk to evaluate the party closely when they check in (training helps them know what to look for). In some cases, such as where a known trafficker has made a live booking, staff could be advised to call law enforcement immediately.
Other factors, which might be made as requests in the booking or at check-in, include requesting specific rooms that present a lower risk for traffickers. Rooms close to an exterior door, for example, can make it easier to let clients in and out undetected. Rooms that are not visible from security cameras, or ones with a view of the parking lot in the direction from which police would arrive, may be requested to reduce risk. Connecting rooms are often used, with the trafficker using one and the victim and clients the other. A property management system could consider these factors whenever a manual room assignment override is requested and pop up an alert to the check-in agent. Requests to skip housekeeping on multi-day stays are also common with traffickers.
Other hotel systems can detect device usage patterns common to traffickers. Some guest room management and locking systems record how often and when the guest room door is opened; trafficking operations often have distinctive patterns. Electronic do-not-disturb indicators can be monitored for constant use. Most guests adjust thermostat settings only once per stay, or perhaps once at bedtime and again in the morning. Traffickers, on the other hand, may reset them for each client, several times in a few hours. Occupancy sensors, particularly in combination with door switches, can detect patterns where one person remains in the room while others are entering and leaving. Guest request apps can track unusual requests such as an excessive number of towels or linens. None of these provide proof of trafficking (they are all things any guest might do), but where enough anomalies are detected, they can help identify rooms where trafficking is more likely, so that hotels can investigate further.
Law enforcement often relies on security camera recordings in investigations and prosecutions. However, many hotel systems fall short because of their age. To be useful, camera resolution needs to be high enough to read license plates in the parking lot or to identify faces in the building. High resolution is state-of-the-art and inexpensive today. The five Ring cameras I have in my house do a better job than many 10-year-old (or older) hotel video security systems, and they cost me only a few hundred dollars. Options with facial recognition have also come down dramatically in cost; they can be used to train cameras on persons of interest.
Signals Detected by Hotel Staff
Many times, an alert hotel staff member sees or hears something that does not seem right. Too often when this happens, they worry about reporting it because it is only a suspicion, and they fear they may get in trouble if they are wrong. Staff need a way to report such things safely, meaning anonymously (if they wish) and in a way that will not call attention to their role.
With the technology used in many hotels today, this should not be a heavy lift. Housekeepers using an app could have a button to report specific (or even vague) things they observed in a particular guest room that might be suspicious: multiple towel or linen sets, unusual trash, lingerie, multiple girls with one man, or even just “something doesn’t seem right.” This could be used to alert security systems or staff to watch the room more closely. Some hotels have trained housekeepers to use their panic buttons to summon supervisory or security staff in situations where they believe there may be a real problem.
Front desk staff could record details about guests checking in, including how many and which guests present themselves at the front desk, whether they have luggage, check-in time, housekeeping requests, and even the demeanor of older children (trafficking victims, if present, will typically avoid any attempt at engagement, even eye contact). This could be prompted by the property management system based on a risk analysis of the booking, to avoid adding too much time to the check-in process.
Front desk staff, engineers, bellmen, valet parkers, bartenders, waitstaff, and room service attendants may also observe suspicious activity. As before, this does not necessarily mean trafficking, but systems can collect and correlate suspicions recorded by human staff just as easily as they can digital activities recorded by door locks, cameras, or thermostats. Hotels simply need to give staff a simple, fast, no-risk way to report concerns. The system can sort out whether, in combination with other data about the same guest or room, it should be cause for concern. A key requirement is to record the exact time of each report, as this can (if necessary) help law enforcement correlate with security video in guest hallways, parking lots, and common areas.
External data sources can also be ingested and used to identify potential trafficking situations. “BOLO” (Be On the LOokout) lists can be compared to guest names in the property management systems. Amber alerts can be broadcast to front desk stations (G6 Hospitality implemented this earlier this year with their property management system provider, HotelKey). This keeps awareness top-of-mind for check-in agents and also provides information about and photos of the victim and, in many cases, the person suspected of trafficking them, aiding identification.
Many traffickers use “burner” cell phones so they can publish a phone number to arrange liaisons. Depending on local laws and phone company cooperation, law enforcement can sometimes track the location of those phones and alert hotels that one appears to be on their premises – a clear signal of a likely active sex trafficking situation. In conjunction with other data and analytics described above, the hotel may be able to help law enforcement identify the guest and room involved.
One hotel chain reportedly identified sex traffickers through shared notes entered by staff in guest profiles. One property first noted suspicious behavior, and others saw it as well and added to it. I would be cautious about doing this in jurisdictions with strict data privacy laws, but the same concept could potentially be done in a privacy-preserving way by aggregating the kind of analytical approach described above across multiple hotels within a brand – if the same guest gets a high risk score at multiple hotels, there is cause for concern.
Analytics are key to assessing the risk that a guest may be trafficking, and the analytics available today are quite limited. PwC now offers a Proactive Guest Safety Monitoring solution (click this link and then scroll down to the mid-page menu and select Cyber, Risk & Regulation). According to the PwC website, it “offers an early warning system that monitors human trafficking, equipping hotel groups with visibility into the risk within their portfolio by supplementing the technology with an action framework for specialists to follow up based on the data alerts.” The emergence of solutions like this, however limited in scope they are today, is essential to capturing and analyzing data and to prioritizing actions.
PwC has a substantial IoT (Internet of Things) practice, which facilitates the connection of signals from multiple sources and makes them a logical candidate to develop such a solution. But many other vendors in the hospitality space have similar infrastructures and could launch similar ones.
Analytics can also be used at the brand, management, or ownership level to compare trafficking-related metrics across a portfolio of hotels. This could measure compliance with training requirements; integration of signal-producing systems; incorporation of facilities in staff applications to report suspicions, adoption and testing of IT policies that can support law enforcement investigations; and guest reviews that may indicate poor overall security generally (e.g., reports of drug deals) or trafficking specifically. Post-departure surveys could add questions about perceptions of the hotel’s overall security.
Many nonprofit organizations are involved in the fight against sex trafficking, and some of them, like Tech Coalition and Enough Is Enough, have collected data that could help inform the analytics described here.
The emergence of analytics platforms tuned to trafficking opens up another opportunity, which is to develop an industry database of trafficking behaviors. Much of the data available today that describes these behaviors is collected ad hoc, often from interviews with hotel staff or survivors. Development of a “trafficking risk score” today requires a lot of assumptions and guesswork based on sparse information. A database that accumulates all available hotel-related data for every reported trafficking incident could greatly improve the ability of systems to accurately identify likely traffickers. While it will be impractical to collect complete and consistent information given the large differences in installed technologies in hotels, we could do far better than we can today.
The use of technology to promote safe environments in hotels is still in its infancy. Many of the same techniques that can be used to help prevent, detect, and ultimately prosecute sex traffickers rely on very simple technologies that can significantly improve safety not only for trafficking victims, but for all guests. It is time for the hotel industry to get the hotels and vendors working together to implement some of the simple changes. It would be great to see one of the major associations serving our industry take a leadership role in making this happen.