Read, Debate: Engage.

Can AI help improve migration management?

December 10, 2023
topic:Refugees and Asylum
tags:#AI, #migration, #refugees, #Syria
located:Syria, Greece
by:Djamilia Prange de Oliveira
When it comes to migration management, AI is often associated with surveillance. But the technology can serve as a positive force for change.  

Artificial intelligence (AI) has been utilised in migration management in the European Union (EU) since the late 1990s, but the use of the technology has continuously expanded over the last decade. The EU commission invests huge sums in the development of lie detector software, dialect identification software, real-time behavioral analysis or data extraction from social media to screen migrants before they enter the EU. 

AI in migration: A bane or boon for migrants?

AI, more often than not, has been detrimental to the civil liberties of migrants and asylum seekers rather than acting as a safeguard. Confronted with lie detectors, visa algorithms and automated border surveillance systems, this already vulnerable and marginalised demographic routinely experiences AI-driven violations of their rights, said Petra Molnar, a human rights lawyer and author of a report entitled Technological Testing Grounds: Migration Management Experiments and Reflections from the Ground Up.

But the technology is not inherently bad, and has the potential to connect people who otherwise wouldn't have met, offering an efficient means to address systemic issues in migration management.

Many of the challenges migrants face, including unemployment, language barriers or mental health issues, often result from a lack of information, network and adequate assistance. However, experts and advocates emphasise that with the right individuals involved, these issues can be resolved.

"If everybody who has the resources to support people would do so every day, we would be able to tackle the refugee crisis much faster," said human rights lawyer Yuliia Karnas at the Arts & Nature Social Club (ANSC) salon in Berlin hosted by ANSC’s programme chair Jörg Geier.

At ANSC, Karnas introduced StartUpBoat, a mobile incubator that generates tech-based humanitarian solutions for refugees on rescue boats in Greece and Turkey.

StartUpBoat was established in 2015 by entrepreneur Paula Schwarz as a social innovation hub that connects refugees with professionals from diverse industries. Fed up with the bureaucracy and inefficiency of public funding for social projects, Schwarz decided to tap the startup world.

"I think Startups can offer great value for humanitarian aid through their technology," she said

Schwarz invited representatives of NGOs, investors, designers, founders, thought leaders, consulters and refugees on boats in the Mediterranean to develop tech-based solutions for the refugee crisis. On one of the first boat rides, the team developed a first-contact website designed to provide information for refugees who had just arrived at the Greek island of Samos. Within a few days, the website had several hundred thousand users.  

Connecting migrants with helpers

Within a few years, StartUpBoat grew into a platform that creates various tech-based solutions in the field of migration, developing projects in healthcare, fintech, connectivity and crisis management that have been used by more than 9 million people.

Its most recent projects are Cosmopolis, a social network app that connects migrants and supporters based on AI and relational data and RefugAI, a free chatbot that assists with translation, bureaucracy and legal advocacy. 

Migrants often struggle to secure employment or access psychological and legal assistance due to language barriers, bureaucratic hurdles or a lack of connections. Cosmopolis aims to address these challenges by enabling newcomers to connect with helpers who can offer legal support, psychological consultations or assistance in finding employment in their new country.

Similarly, with RefugAI, refugees and asylum seekers can overcome bureaucratic obstacles and language barriers by having the AI break down complex information and translating it.

ANSC’s programme chair Jörg Geier added that such comprehensive approaches are essential to understand the situation on the ground and tackle the root causes of the challenges at hand.

Transforming humanitarian aid through AI

According to social entrepreneur Ummul Choudhury, AI can also transform the sector of humanitarian aid, which struggles with complex bureaucratic hurdles.

Choudhury founded Capoeira4refugees, a sports-charity that provides support to traumatised refugee children through the Brazilian martial arts Capoeira, in Damascus, Syria. But while managing the charity, she found herself overwhelmed by systemic and bureaucratic issues.

"Only 2 per cent of international aid actually makes it to local organizations," Choudhury said at the Arts & Nature Social Club salon in Berlin. Most of the money, according to Choudhury, is stuck in the hands of larger organisations or lost in a chain of international transfers between larger and smaller NGOs. 

The distribution of aid funds involves major organisations like the UN, the EU, or USAID subcontracting smaller NGOs. But these local entities often face challenges meeting the stringent requirements imposed by donors due to limited capacity and resources. Meanwhile, larger international organisations incur significant administrative and logistical costs. Compounding these issues, organisations operating in conflict zones may encounter obstacles such as limited access to the banking system and a lack of basic infrastructure.

"By the time any of the money gets to a Syrian organisation, there is only 2 per cent left," Choudhury said. "And the reason why nobody wants to change that is because the system doesn’t trust local people to handle the funds. But tech can change that." 

"Getting the money to the right people is about technological solutions"

"If data is transparent, and we can ensure that money isn’t going to criminal activities, terrorism or militias, we can create trust through technology and thereby overcome bureaucratic hurdles – and that is what we are doing," Choudhury added.

To ensure aid funds reach their intended destination - the local population - she established the real-time data collection app FrontlineAid and the charity RealtimeAid. Both initiatives collaborate with local organisations in Syria.

FrontlineAid collects information from grassroots organisations in need of funding, including videos, voice memos and text-based data. Operating as both an onboarding and project management tool, the platform enables local organisations to share details about their projects and determine where financial support is most required.

 Projects that pass the initial onboarding process are assigned to a mentor who works with them to develop an action plan and meet donor requirements.

After localising aid money, RealtimeAid, in partnership with FrontlineAID, handles the necessary paperwork for the local organisations to receive funding. The information stored in the app is visible to both local NGOs and donors, allowing organisations to easily generate transparent reports about their operations and financials. 

RealtimeAid currently supports community projects of 10 local organisations in Al Raqqa, Syria, ranging from learning centres for special needs children, sports classes for traumatised youths and refurbishing former ISIS prisons into safe play spaces for youth and women.

Fostering transformative change 

But Choudhury has also encountered challenges in the process of deploying technological solutions in humanitarian aid. In addition to navigating diverse cultural norms and local hierarchies in Syria, she reflects on the difficulties of implementing an online tool in a context where command of such platforms was minimal among the targeted group, as detailed in a case study.

Moreover, reliance on AI poses the risk of in-built bias, a challenge commonly observed in language models like ChatGPT or Bard. Choudhury emphasises the importance of asking whether we want an algorithm to reflect the world as it is or as we believe it should be.

"As a human interpreter, you have to understand that an algorithm trained on a certain pool of data by a certain pool of people will reflect that data, and it will therefore always be somewhat biased," she shared. "But if you know that, you can work around it. As organisations like us take on AI, we have to educate ourselves." 

According to Choudhury, the human element cannot be taken out of the equation when it comes to tackling systemic issues in migration management.

Image by Julie Ricard.

Arts & Nature Social Club, co-chaired by Jörg Geier, is a community of opinion leaders in the fields of arts, science and business that hosts events to drive sustainable change.

Article written by:
Böll Moderation
Djamilia Prange de Oliveira
Author
Syria Greece
Embed from Getty Images
Confronted with lie detectors, visa algorithms and automated border surveillance systems, this already vulnerable and marginalised demographic routinely experiences AI-driven violations of their rights.
Embed from Getty Images
AI has the potential to connect people who otherwise wouldn't have met, offering an efficient means to address systemic issues in migration management.
Embed from Getty Images
Cosmopolis is a social network app that connects migrants and supporters based on AI and relational data. RefugAI is a free chatbot that assists with translation, bureaucracy and legal advocacy. 
.
.