Meet Darb Alaman: The Chatbot Fighting Domestic Violence In The Middle East 

TW: This article contains mentions of domestic violence and sexual abuse. 

At the edge of the Mafraq desert in northern Jordan, where laundry flaps in the wind outside sun-bleached cinderblock homes and silence can be a survival strategy, something quietly radical is unfolding. It’s not a mass protest or a new law passed in the capital, Amman. Instead, it lives in WhatsApp chats and Facebook Messenger threads, often opened in secret, sometimes at 2am, by women like Arwa*.

When the beatings got worse, Arwa waited until her husband left the house. Then she opened Messenger and began to type. Her message was short: “I need help. He hurts me. My children are afraid.” At the other end of the line wasn’t a friend or family member, but a discreet, human-led chatbot: Darb Alaman. With a name that means ‘Safe Pathway’ in Arabic, Darb Alaman – launched by ActionAid Arab Region (AAAR) – lives in messenger apps, designed to support survivors of domestic and gender-based violence in a country where speaking out can carry devastating consequences. 

In Jordan, for far too many women, violence isn’t an isolated incident – it’s the backdrop of everyday life. Around one in four has experienced domestic abuse, but accessing help can be a minefield. In fact, it can be dangerous. The country lacks a standalone law to address domestic violence and the protections that do exist are fragmented at best – often funnelled through male-dominated institutions that can be more punitive than protective. For displaced communities – and Jordan hosts the second highest number of refugees per capita in the world, after Lebanon – the danger only deepens. Speaking out can cost women their homes, their children, or even their right to remain in the country.

This is the context in which something like a chatbot can make a life-changing difference. Because in Jordan, asking for help is not as simple as walking into a police station. “Women are terrified their families will find out. That they’ll lose their children. That they’ll be killed.” says Farah Mustafa, a project officer with AAAR.

The statistics are jarring. A 2017-18 national survey found over a quarter of Jordanian women had experienced domestic violence at the hands of their husbands. For Syrian refugee women, the rates are higher: more than 30% report physical abuse and more than half have endured emotional violence. In 2023, reported domestic violence cases jumped by 38% compared to the previous year, reaching over 58,000. Last year, of 12 family-related homicides, nine of the victims were women. And that’s only what’s been reported.

For this very reason, Darb Alaman is discreet. When a woman opens the chat on WhatsApp or Messenger, a quiet system flickers to life. No fanfare, no bright logos. Just a simple confidentiality notice, a choice between Arabic or English and an interface designed to blend in. It’s not meant to dazzle. It’s meant to protect.

From there, a series of calm, straightforward prompts: country, gender, age group, nationality. Then the core question: what does she need? Information, or help? If she taps ‘general information’, she’s taken through a menu of subjects most people prefer not to talk about: rape, forced marriage, cyberbullying, childhood abuse. It’s not sugar-coated, but it’s clear, clinical. A kind of lifeline in bullet-point form.

But if she’s in need of more than knowledge – if she’s in danger, or trying to escape – a woman can choose ‘service providers’. Behind that button lies a carefully vetted list of legal aid groups, therapists, health clinics, protection services and even education contacts. All localised, all tested.

Farah, who has been involved in the project’s development for three and a half years – following the bot’s launch a year prior – has spent years visiting homes across Jordan’s provinces. She speaks bluntly about the realities women face: early marriage, lack of education, economic dependence and an overwhelming stigma around divorce and mental health. She points to a specific challenge rooted in Jordan’s Nationality Law No. 6 of 1954, which doesn’t allow Jordanian women to pass their nationality to their children or non-Jordanian spouses. This creates real barriers for women who wish to marry outside their nationality and adds to a broader sense of restriction.

“The country is trying, but most women still feel they don’t have the right to speak,” she says. “Even therapy is a luxury few can afford. We wanted to create a safe, confidential and accessible way for women to get help anytime, anywhere.”

So now women can now log onto a social media platform they already use and be guided, confidentially, towards support. The entire system is built for women who may not be fluent in apps or technology. The language is clear and compassionate, and it resembles a personal profile, so all users have to do is engage with it as they would a familiar connection – simply send a message to trigger a reply. There’s no chatbot banter, just structured menus and instant responses. If the internet cuts out mid-use, the chat picks up right where it left off once she’s back online. No data lost. No need to restart. No trace left behind.

Crucially, anonymity is baked into its DNA: users are never asked for full names or personal details. Their stories are received not by a machine, but by trained social protection officers embedded within a confidential, vetted system. Farah says: “We could have made it more ‘smart’, but in this context, smart isn’t what people need. They need to feel safe.” She explains: “We don’t collect sensitive data. Everything is confidential. Only trained professionals see what’s shared and not any external companies.”

With confidentiality in mind, Darb Alaman doesn’t ping with follow-up messages nor show up as a notification. And while there’s no emergency exit button yet, the tool’s static, non-conversational design keeps it low-risk. There’s no digital footprint, because in many of the homes this tool was built for, the most dangerous person in the house is watching the phone. There’s also no legal risk to users, or to the organisation. The project is fully compliant with Jordanian law and international data protection standards.

What sets Darb Alaman apart is not only what it does, but how it was built. This wasn’t a top-down innovation shipped in from a foreign capital. Its architecture was shaped through consultations with women who had lived through abuse, particularly from refugee communities. ActionAid and local Women Protection Action Groups spent months gathering stories and testing prototypes. They discovered that 70% of women weren’t aware of their legal rights, 40% didn’t know where or how to access services – or couldn’t leave home to do so – and 65% said they preferred digital platforms for support.

That data shaped everything, and the use of Facebook Messenger was key, Farah explains, as it’s a platform widespread and familiar across the country, even to the most tech-averse users. Data on the bot’s usage from this year so far tells a story of growing traction, with nearly 40,000 digital views and over 2,200 unique visits. Content interactions are up 4,000%, showing that the tool is not just a curiosity, it’s a lifeline. Many of them are young women under 35 – digital natives but still navigating deeply traditional expectations.

Fortunately, the government isn’t getting in the way. In a region where online tools can attract state suspicion, this one hasn’t. Farah explains: “Our work aligns with Jordan’s national strategy on violence against women. We also collaborate with relevant ministries and stakeholders to ensure that our tools are complementary to national protection systems.”

Volunteers and community-based organisations remain key, Farah says. Trained in gender-based violence (GBV) response, they lead awareness sessions across Jordan. Usage of Darb Alaman has steadily risen, however the impact is harder to track. Farah says: “We currently do not have a quantified follow-through rate from digital interactions to physical services. The chatbot is designed to guide users to accessible support services across Jordan.”

But numbers only say so much. The power of Darb Alaman is written in the stories of the women it has reached. Women like Arwa. For her, the chatbot was not just a source of information, but a gateway. Through it, she received trauma counselling and later, a startup kit to launch her own home bakery. “I make money now,” she says. “I feel like I can breathe.”

But building that pathway hasn’t been without difficulty. Some users, like Nouf*, encountered barriers not of bandwidth but literacy. In her sixties, Nouf had never learned to read and was initially bewildered by the chatbot’s interface. But through patient guidance and determination, she found her way not just into the digital world, but into a space of self-discovery. Farah says: “At the age of 65, she learned to read. And she started questioning the norms she’d lived with all her life.”

Still, the limits of Darb Alaman are as important as its successes. As with all digital tools, access is not universal, especially among older women or those in rural areas. Many users are unfamiliar with chatbots altogether. Farah says the team is currently developing a voice-note feature to reach those who cannot read or write, but expanding the tool is slow-going and underfunded.

In fact, the project is nearing the end of its four-year funding cycle. “We’ve kept operations running, through private funding, which has enabled the full development and deployment of the tool,” Farah explains. She adds that Darb Alaman is in its final stage of implementation, focusing on refining the tool, conducting outreach and ensuring it is fully functional and accessible to the target users. “Our budget does not allow for all the upgrades or the expansion to additional countries that we had envisioned,” she says, adding that attempts to scale the chatbot into Lebanon were paused due to logistical hurdles and difficulties keeping referral networks current. She says these gaps are indicative of funding for the “wider GBV ecosystem”.

However, in the global “AI for good” conversation, Darb Alaman offers a rare case study in restraint and community-first design. While some international projects make sweeping claims about AI revolutionising humanitarian response, this tool keeps its ambitions deliberately narrow. It doesn’t try to predict abuse, map risk or automate empathy. What it offers instead is a quiet, culturally grounded alternative: a way to reach out without being seen, a chance to act without asking permission.

Could it be replicated? Yes, but while the technology is transferable, the trust it’s built on is not. Any replication would need to be locally led, deeply contextual and wary of digital saviourism. “We don’t want this to be a tech solution imposed from outside,” says Farah. “We want it to be something communities shape themselves. But whether we like it or not, AI is here so we should use it for good.”

There is, of course, a danger in romanticising chatbots, for they can’t rewrite laws and won’t dismantle gender norms. But they can create small sanctuaries of safety within oppressive systems. For someone like Arwa, typing a message in the middle of the night with a bruised face and terrified children by her side, that sanctuary might be the difference between silence and survival. And in homes across Jordan, that quiet revolution continues, one message at a time.

*Names changed to protect identities