When Human Care Meets AI: Why Vita Activa Decided to Go to Zambia

Conecting people to care

At RightsCon 2026 in Zambia, Vita Activa planned to publicly introduce NINA, our new AI-assisted conversational triage system designed to support journalists and media workers seeking help through MDRnet.org (Media Resilience Network), Vita Activa’s program for the US. The conference was suspended before our launch could happen. Yet we decided to travel and, in many ways, that mattered even more.

By Ana Arriagada
Managing Director – Product and Implementation, Vita Activa

Being there allowed us to meet with organizations, activists, journalists, technologists, and care practitioners from across Africa and beyond. Those conversations reinforced something we already suspected, but could now see more clearly: the urgent need for emotional support infrastructures for people working under extreme pressure, especially journalists, activists, women, and LGBTQ+ communities facing systemic violence and chronic stress.

Across many conversations, one reality kept surfacing: people are exhausted, isolated, and often navigating environments where support systems simply do not exist or remain inaccessible.

That context is deeply connected to why we are building our new platform, NINA (the acronym in Spanish for Our Intelligence is not Artificial – Nuestra Inteligencia No es Artificial).

What is NINA?

NINA is an AI-assisted conversational triage system that helps guide MDRnet users (and, in the future, Vita Activa’s users) toward the support and resources they need in a clear and efficient way. Its role is to handle first contact, understand users’ situations, and orient them toward the most appropriate type of support within MDRnet.

It does not replace people. Its purpose is to facilitate access to human care efficiently 

What users encounter may initially feel familiar. Like systems such as ChatGPT, Gemini, Claude, or Perplexity, NINA uses conversational AI to understand what someone needs and help organize the next steps. But the philosophy behind it is fundamentally different.

NINA was intentionally designed not to simulate care, but to connect people to care.

Why We Built NINA

For seven years, Vita Activa has operated as an emotional first aid helpline. Our work has never been about offering magical solutions or automated reassurance. Instead, we accompany people as they regain agency, navigate difficult situations, and make informed and conscious decisions about what comes next after traumatic events or emotional crises.

In many ways, Vita Activa’s work has always involved co-creating possible paths alongside the people who reach out to us.

That experience shaped the core design questions behind NINA:

• How do we respond to people as quickly as possible, even when no human responder is immediately available?
• How do we avoid losing contact with people who reach out once, but disappear after long waiting times?
• Can AI help organize intake, documentation, and triage in ways that reduce the burden on care teams instead of replacing them?

Another turning point was a crisis we documented in our recent report on AI and care infrastructures. In September 2025, Vita Activa experienced a massive influx of referrals originating from AI systems and automated recommendation flows. The volume overwhelmed our small team and temporarily collapsed parts of our support capacity.

That experience forced us to confront a difficult reality.

If people are already turning to large language models seeking mental health support even when those systems can only simulate empathy

What responsibility do care organizations have in shaping more ethical alternatives?

And if major platforms increasingly redirect high-risk or emotionally distressed users toward human organizations when situations become too complex or legally sensitive, can smaller care infrastructures develop tools that actually prioritize ethics, responsibility, and human context over speed and the appearance of efficiency?

These are some of the questions behind NINA.

Humans Supporting Humans

NINA is only the entry point. Care itself remains fully human.

Journalists using MDRnet can expect a first interaction that is clear, confidential, and structured. NINA asks a few brief questions to determine whether the user is within MDRnet’s scope and what kind of support they are seeking. Based on that, it guides them toward the appropriate next step, usually connecting them with the MDRnet team through Signal, our primary secure communication channel.

The support itself happens between people.

This distinction matters deeply to us.

Machines can simulate conversation. But human life requires more than conversation: it requires care.

When someone is experiencing distress, they need another person capable of listening, understanding context, intervening responsibly, and holding ethical accountability. These are not functions that algorithms can fulfill.

NINA was built with that limitation in mind.

Rather than extending automated interaction indefinitely, it prioritizes timely and responsible connection to human support. It also helps communicate something increasingly difficult in digital culture: human care moves at a different pace than automation.

The expectation of immediacy created by digital platforms has shaped how many people seek support today. But meaningful care cannot always be instant, scalable, or frictionless.

Recognizing those limits is not a failure of care. It is part of what makes care human.

Zambia and the Future of Care Infrastructures

Traveling to Zambia after RightsCon’s suspension reminded us that these conversations cannot remain centered only in the Global North or inside technology spaces.

Many organizations we met are already supporting communities under extraordinary political, economic, and emotional pressure, often with very limited infrastructure and resources. The need for ethical, community-centered support systems is urgent.

NINA is, ultimately, an experiment. One grounded in years of frontline emotional support work, but an experiment nonetheless.

We are open to whatever conclusions emerge from that process. The answer may ultimately be that AI can responsibly support certain parts of care infrastructures. Or it may reveal limits that should not be crossed.

What matters to us is remaining honest about those limits, transparent about the risks, and committed to protecting the human relationships at the center of care.

Because technology may help organize support. But it cannot replace the people who sustain it.

---

---

If you are a US based journalist, reach out to MDRNET.Org