Can AI be a tool to advance Positive Safety?

The Shifting Meaning of Safety

From there, thinkers like Locke, Bentham, and Mill tied security to the state’s role in maintaining order, cementing its modern association with law enforcement and threat prevention (Zedner, 2009). Today, much of safety discourse remains trapped in this Hobbesian mindset—framed as the absence of danger rather than the presence of well-being.

The Problem with "Negative Safety"

For decades, security and safety has been viewed through a lens of suspicion and control. Scholars critique "securitization"—the process of framing issues as existential threats to justify extreme measures (Bigo, 2002; Buzan et al., 1998). This approach thrives on binaries: us vs. them, safe vs. unsafe, order vs. chaos.

But this mindset has limits:

• It assumes humans are inherently aggressive, requiring top-down control.
• It fuels fear-based policies, from mass surveillance to predictive policing.
• Despite declining crime rates, people feel less secure (Schuilenburg & van Steden, 2014)

Even AI-powered tools—facial recognition, predictive algorithms—are often deployed in ways that reinforce this negative paradigm, focusing on "combating," "controlling," and "punishing" (Schuilenburg et al., 2017). But what if safety could be about more than just preventing harm?

Introducing Positive Safety

A growing movement of scholars is flipping the script. Instead of asking, "How do we eliminate threats?" they ask, "What conditions enable people to feel safe?"

Positive Safety reimagines safety as:

• A dynamic relationship—not just state protection, but mutual care (McSweeney, 1999)
• Rooted in trust and belonging—prioritizing social cohesion over surveillance (Schuilenburg & van Steden, 2014).
• Empowering communities—bottom-up solutions that leverage local networks (Gjørv, 2012)

This shift aligns with the UN’s human security framework, emphasizing the root conditions of well-being (Paris, 2001)

The Untapped Potential of AI for Positive Safety

Most AI applications in public safety today reinforce old paradigms—predicting crime, monitoring crowds, flagging "risky" individuals. But what if AI could instead:

• Strengthen community trust?
• Enhance social support networks?
• Foster inclusive participation in safety measures?

At AI-MAPS, we’re exploring exactly that. How do you envision "positive safety"?

Sources

Arends, J. F. M. (2018). ‘From Homer to Hobbes and Beyond – Aspects of “Security” in the European Tradition’. In H. G. Brauch, Globalization and Environmental Challenges. Reconceptualizing Security in the 21st Century (pp. 263–277). Springer.

Bigo, D. (2002). Security and Immigration: Toward a Critique of the Governmentality of Unease. Alternatives: Global, Local, Political, 27(1_suppl), 63–92. https://doi.org/10.1177/03043754020270S105

Buzan, B., Wæver, O., & De Wilde, J. (1998). Security: A new framework for analysis. Lynne Rienner Publishers.

Gjørv. (2012). Security by any other name: Negative security, positive security, and a multi-actor security approach. Review of International Studies, 38(4), 835–849.

McSweeney, B. (1999). Security, identity, and interests: A sociology of international relations. Cambridge University Press.

Paris, R. (2001). Human Security: Paradigm Shift or Hot Air? International Security, 26(2), 87–102. https://doi.org/10.1162/016228801753191141

Schuilenburg, M., Schoenmakers, Y. M. M., & Zanten, J. (2017). Positieve veiligheid. Naar een nieuwe Rotterdamse veiligheidsstrategie. VU Amsterdam.

Schuilenburg, & van Steden, R. (2014). Positive security: A theoretical framework. In Positive Criminology: Reflections on Care, Belonging and Security. Eleven.

Zedner, L. (2009). Security. Routledge. https://doi.org/10.4324/9780203871133