Center for International Policy
LightNews LightNews
banner
cipolicy.bsky.social
Center for International Policy
@cipolicy.bsky.social
2.4K followers 40 following 220 posts
Advancing a more peaceful, just & sustainable world by centering people & the planet in US foreign policy.

Home of International Policy Journal and UnDiplomatic Pod, The Iran Podcast and Black Diplomats.

Learn more at internationalpolicy.org
Posts Replies Media Videos
Reposted by Center for International Policy
atherton.bsky.social
This remains the most prescient piece I've ever published at @cipolicy.bsky.social internationalpolicy.org/publications...
Trump’s Wag The Dog Moment  - CIP
If you want to break up a drug cartel, blowing up a boat does little to accomplish that, unless you believe cartels only own one boat.
internationalpolicy.org
November 23, 2025 at 9:48 PM
cipolicy.bsky.social
If you like this and other policy-forward writing, you can get our International Policy Journal newsletter in your inbox open.substack.com/pub/cippolic...
Can Complimentary Learning Methods Teach AI the Laws of War?
BASIC Training, Doubt Protocols
open.substack.com
November 10, 2025 at 5:26 PM
cipolicy.bsky.social
This argument is compatible and complimentary with a strong arms-control approach to the development and fielding of military AI, carving out a space of what more-responsible AI development looks like, to contrast with how military AI is at present. internationalpolicy.org/publications...
Can Complimentary Learning Methods Teach AI the Laws of War? - CIP
Training AI models on edge cases of war offers the possibility of hard-coding compliance with laws or war, or at least necessary doubts.
internationalpolicy.org
November 10, 2025 at 5:26 PM
cipolicy.bsky.social
Davit Khachatryan is thorough on what coding for international humanitarian law will require, and the piece offers concrete recommendations for how to mandate accountability in military AI, rather than using it as a shield to obfuscate responsibility.
SOLUTIONS, DISTILLED: Can Complimentary Learning Methods Teach Al the Laws of War? • Balance imitation and reinforcement learning in military Al training, Il to inherit human judgment patterns, RL to adapt to novel scenarios, while correcting both methods' statistical bias against rare but critical restraint decisions. • Design simulations to overrepresent improbable, high-stakes scenarios so that rare legal and humanitarian challenges carry statistical weight in the model's decision-making. • Embed the humane fallback into Al logic by oversampling lawful restraint in Il datasets and shaping RL rewards to prioritize caution under uncertainty. • Install non-overridable constraint layers and doubt protocols so that when target identification is uncertain, lethal action defaults to suspension or human review. • Conduct counterfactual and adversarial testing to verify that humane tallback behaviors persist under deceptive inputs, sensor errors, and operational stress. Davit Khachatryan for the International Policy Journal
November 10, 2025 at 5:26 PM
cipolicy.bsky.social
We can mandate higher standards for such AI use.

"Embedding the humane minimum into AI means that in every training run, whether through curated historical cases or artificially generated edge scenarios, the option that aligns with humane treatment under uncertainty must be given decisive weight."
Can Complimentary Learning Methods Teach AI the Laws of War? - CIP
Training AI models on edge cases of war offers the possibility of hard-coding compliance with laws or war, or at least necessary doubts.
internationalpolicy.org
November 10, 2025 at 5:26 PM
cipolicy.bsky.social
"The way those objectives are weighted in the reward function is decisive. If mission success is rewarded heavily and civilian harm only lightly penalized, the AI will statistically favor the course of action that maximizes mission success, even if that means accepting higher risks to civilians."
Can Complimentary Learning Methods Teach AI the Laws of War? - CIP
Training AI models on edge cases of war offers the possibility of hard-coding compliance with laws or war, or at least necessary doubts.
internationalpolicy.org
November 10, 2025 at 5:26 PM
cipolicy.bsky.social
"Left uncorrected, the AI may treat those lawful restraint decisions as statistical noise, unlikely to be repeated in practice."

internationalpolicy.org/publications...
Can Complimentary Learning Methods Teach AI the Laws of War? - CIP
Training AI models on edge cases of war offers the possibility of hard-coding compliance with laws or war, or at least necessary doubts.
internationalpolicy.org
November 10, 2025 at 5:26 PM
cipolicy.bsky.social
Training machines to hesistate and hold fire when the laws of war demand it might be more a question of will than tech. As Davit Khachatryan argues, both imitation and reinforcement learning could oversample situations that mandate restraint.

internationalpolicy.org/publications...
Can Complimentary Learning Methods Teach AI the Laws of War? - CIP
Training AI models on edge cases of war offers the possibility of hard-coding compliance with laws or war, or at least necessary doubts.
internationalpolicy.org
November 10, 2025 at 5:26 PM
Reposted by Center for International Policy
atherton.bsky.social
Really happy to publish @wesinjapan.bsky.social at @cipolicy.bsky.social on the history and future of disaster response internationalpolicy.org/publications...
DOGE’s Gutting Of USAID Weakened Disaster Response - CIP
For over 30 years, USAID had a leading role is disaster response for the United States, a capacity shattered when it was shut down.
internationalpolicy.org
November 4, 2025 at 9:03 PM
cipolicy.bsky.social
@w2t2impact.bsky.social reveals how the 1122 Program funneled over $100 million in military-style equipment and surveillance software to law enforcement at federal discounts, with minimal transparency or accountability.🔻

static1.squarespace.com/static/66269...
static1.squarespace.com
October 30, 2025 at 4:27 PM