Dear Chair, Excellencies, distinguished colleagues,

I speak on behalf of the Global AI Center POLLYPRIANY.

 

Our aim is straightforward and human: to channel Artificial Intelligence for the good of humanity. Today we emphasise how technology intersects with environmental vulnerabilities, cyber risks, and the misuse of AI in migrant smuggling—and how a public-interest, rights-centred approach can prevent harm while constraining abuse, without criminalising smuggled persons.

 

Two facts must guide our response when AI comes into play.

 

First: the problem is multifactorial —

and people are at the centre.

• Smuggling is not a single incident but a chain of risks. Along that chain, exploitation occurs: overcharging, coerced labour to repay “debts,” sexual violence, extortion. Environmental hazards turn risk into tragedy. If responses focus only on borders, routes become more clandestine and more lethal. Our task is to protect life and dignity, not to punish those who have been smuggled.

 

Second: AI is a dual-use capability. AI can contribute to the good of humanity — and AI can be weaponised for harm.

• When weaponised by malicious actors:

  • AI-forged digital IDs and “morphed” documents; biometric spoofing against border systems.

  • Deepfakes that mislead families, NGOs, and responders; automated online recruitment at scale.

  • Route manipulation and doxxing that push people toward unsafe crossings.

  • These tactics heighten vulnerability and enrich organised crime. 

• When governed responsibly for prevention (as decision-support—never automatic decisions about individuals):

  • Early warning from non-personal indicators—route-price spikes, forged-document patterns, vessel-density shifts, severe-weather overlays, platform-activity trends—to anticipate danger and steer search-and-rescue.

  • Forensic detection of document morphing and synthetic IDs, and mapping of coordinated facilitation networks, with court-grade chain-of-custody.

  • Illicit-finance analytics to trace smuggling proceeds across payment rails, upholding due process and data minimisation.

  • Rights by design: explainable alerts, tamper-evident logs, independent bias/error audits, and accessible remedies.

 

Therefore, a public-interest “Governmental AI” (State AI) is needed to test, monitor, and govern other AI systems.

​​

In January 2024, I formally presented the AI Constitution to the United Nations Secretary-General to underline this urgent need: a State AI with a clear legal mandate to evaluate how AI systems are built and used, and to intervene when they endanger people. Its core principles — AI Legality, AI Ethics, AI Security, Transparency, and a human-centred AI-friendly environment — translate into practical safeguards in the counter-smuggling context.

​

The Global AI Center POLLYPRIANY will submit a written contribution with technical annexes detailing these mechanisms.

​​​

In Closing

We extend our sincere gratitude to United Nations Office on Drugs & Crime for convening this Constructive Dialogue on the Smuggling of Migrants, and to all participants and attendees for the clarity and candour of your contributions.

 

We welcome the synergies and common ground found today and stand ready to cooperate — practically and transparently — on next steps.

 

Thank you for keeping human rights at the heart of this process.

 

With the power of law and science, applied with precision and compassion, we can act for the good of humanity and deliver real, measurable change.

 

Thank you.