The dark future of AI in war | Stephanie Hare | TEDxBerlin

Item: The dark future of AI in war | Stephanie Hare | TEDxBerlin
Rating: 4
Author: PostSingularity Institute

by Stephanie Hare · View the channel on PSI

Humanities, Social Sciences & Thought AI & Society UBDigital and Information technology: general topics UBJDigital and information technologies: social and ethical aspects

🎙 Stephanie Hare 👥 44.6M 📅 September 18, 2026 ⏱ 14 min 👁 639 📄 expert opinion 🧭 2026-09-18

Available in: English (current) Français

Keywords

AIwaraccountabilityethicsautonomous weapons

Summary

Stephanie Hare, author of ‘Technology Is Not Neutral’, discusses the ethical and political challenges of AI in warfare. She contrasts the optimistic ‘use AI or be left behind’ narrative with warnings from AI leaders about existential risks, reframing the question from ‘who is right?’ to ‘how do we ensure human accountability?’. She introduces lethal autonomous weapon systems (LAWS) that operate without human control, noting they are designed to kill and that current AI unreliability risks civilian casualties. She examines a public clash between US President Trump and Anthropic CEO Dario Amodei: Amodei sees reliability as the main technological hurdle, while Trump insists on presidential authority to use such weapons. She also cites a UK minister’s contradictory statement about removing humans from the loop. Hare argues that technology is never neutral—humans are always in the loop, and those who finance, build, sell, use, or fail to regulate these weapons can be held accountable under existing law. She proposes three options: outright ban, deterrence (mutually assured destruction), or continued inaction, and points to the Nuremberg trials as precedent for holding human leaders and executives responsible for war crimes. She concludes that AI cannot be tried, but humans can and should be.

199 words

Critical Evaluation

Value of the Information & Strength of the Argument

The talk delivers strong value by reframing a complex technological and ethical issue into a clear question of human accountability. Stephanie Hare’s argument is logically structured: she acknowledges the debate between technological feasibility and political authority, then pivots to the under-discussed aspect of legal responsibility. Her use of the Nuremberg precedent is powerful, showing that ‘just following orders’ is not a defense, and she extends this to AI decision-making. The argumentation is compelling because it appeals to existing legal frameworks rather than proposing novel mechanisms, making the case practical. However, the talk is primarily an opinion piece supported by contemporary events, and while it cites specific facts (e.g., Pope’s encyclical, UK minister’s statement), it does not provide direct citations, relying on the audience’s trust in the speaker’s expertise.

Scientific Rigor, Source Quality, Title Accuracy

The scientific rigor is moderate: she references credible sources (her own book, Pope’s encyclical, statements from public figures) but does not provide formal citations within the talk. The quality of sources is acceptable for a TEDx presentation, given the speaker’s background and the use of widely known events. The title is highly accurate—the ‘dark future’ is central throughout, and she explicitly discusses the dangers of autonomous weapons. The talk is well-structured, moving from the broader AI discourse to the specific case of lethal autonomous weapons, and ends with a call to action. No audience comments were provided, so no public sentiment analysis is included.

247 words

Title / Content Match

The title accurately reflects the content: the talk directly addresses the dark implications of AI in warfare and the challenges of accountability.

Quality & Reliability

8/10

The speaker is an expert with a PhD, published author, and credible background. She presents a clear argument with real-world examples, though some claims are not formally sourced within the talk.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Cited Sources

TEDx Talks — Official page for TEDx events, described in the video description.

Contribution & Novelties

The talk contributes an accessible and compelling argument for maintaining human accountability over AI in warfare, synthesizing recent political and industry events with historical legal precedent. It challenges both the ’technological progress’ and ‘political authority’ framings, advocating for a legal accountability lens. The originality lies in connecting contemporary debates to the Nuremberg trials, emphasizing that existing international law is sufficient to hold individuals responsible.

Pour aller plus loin :

Lethal autonomous weapon — Overview of LAWS and associated debates.
Human in the loop — Definition and significance in automated systems.
Nuremberg trials — Historical context for the principle that individuals are responsible for war crimes.
Technology Is Not Neutral — Speaker’s book, though not explicitly referenced with a URL in the talk, it is a direct source.

127 words

Radar Profile

The radar profile shows high quality information (8/10) and moderate quantity (7/10), indicating a well-argued but not overly dense talk. Technical level is low (5/10), making it accessible to a general audience. Overall reliability is solid (7/10), reflecting the speaker's expertise, though some assertions lack direct citation.

Reliability 7/10