The online home of the Berkeley Risk and Security Lab @UCBerkeley from @GoldmanSchool and @BerkeleyISchool; technology, politics, security
Berkeley, CA
Joined July 2022
- Tweets323
- Following255
- Followers594
- Likes450
Berkeley Risk and Security Lab retweeted
I'll be speaking at "Algorithmic Warfare in Iran and Beyond" on April 1st. I'm joined by the fantastic @jessicadorsey , @mchorowitz, and @Ritwik_G. Please bring your questions!
Big thanks to @BerkeleyRisk for hosting us! RSVP below.
berkeley.zoom.us/webinar/reg…
Berkeley Risk and Security Lab retweeted
Next Monday, October 27, UC Berkeley Risk and Security Lab will be hosting MTA Senior Researcher Dr. Hui Zhang for a discussion on his forthcoming book "The Untold Story of China’s Nuclear Weapon Development and Testing"
brsl.berkeley.edu/beyond-the…
Berkeley Risk and Security Lab retweeted
⏰ HAPPENING TODAY
How will AI agents shape peace and security?
Join SIPRI’s webinar today at 17:00 CEST to find out.
Read more🔗 bit.ly/4nYLJls
Register now🔗 bit.ly/3Kymlog
🎙️ @_achan96_, @GovAI_; @jzl86, @GoogleDeepMind; @SarahShoker, @BerkeleyRisk
Moderator: @BoulaninSIPRI
Berkeley Risk and Security Lab retweeted
How can we develop #AI to ensure human flourishing? 🤖
Join Stuart Russell at @UCBerkeley #TechPolicyWeek as he explores the future of AI.
🗓️ Oct 21 | 4:30–6pm | 📍 Sutardja Dai Hall
🔗 eventbrite.com/e/ai-whats-to…
Co-hosted by @BerkeleyCDSS, @BerkeleyKCESP & @BerkeleyRisk
Berkeley Risk and Security Lab retweeted
How will AI agents shape peace and security?
🗓️ Join SIPRI’s public webinar on 15 Oct, 17:00 CEST to find out.
Read more🔗 bit.ly/4nYLJls
Register now🔗 bit.ly/3Kymlog
🎙️ @_achan96_, @GovAI_; @jzl86, @GoogleDeepMind; @SarahShoker, @BerkeleyRisk
Moderator: @BoulaninSIPRI, #SIPRI
#AI #Security #PeaceTech #AgenticAI
Berkeley Risk and Security Lab retweeted
Join @BerkeleyRisk and @CLTCBerkeley for a panel discussion of Dr. Elaine Korzak's new paper "Managing Commercial Spyware Through Export Controls"!
🗓️June 18th, 10-11AM PT
🌐Zoom Webinar
✅Register to attend here: cltc.berkeley.edu/event/mana…
Berkeley Risk and Security Lab retweeted
How the government responds to the competition between the US and #China will be one of the most important challenges for the new administration. We need to get it right...
lawfaremedia.org/article/the…
Ritwik Gupta and @areddie89 explain what is in Senator Hawley's new AI bill aimed to stop collaboration of AI tech between the U.S. and China and why it will stifle American AI progress and innovation.
Berkeley Risk and Security Lab retweeted
Instead, we suggest that data itself might serve as a place to center governance efforts.
Well done to the @BerkeleyRisk and @berkeley_ai teams on this--and particular thanks to @Ritwik_G and @LPWalkr for leading the charge!
Berkeley Risk and Security Lab retweeted
As anyone who has been following various #AI #governance debates knows, a great deal of ink has been spilled on whether and how to think about limiting the size of these models.
But is this the right approach? In our latest, we suggest not.
arxiv.org/abs/2409.17216
Berkeley Risk and Security Lab retweeted
Just re-upping this thread with 3 working days to go until the first review date on August 7. If you're interested in working at the intersection of #technology #politics and #security, join us!
Big news from Berkeley Risk and Security Lab. We are hiring for FOUR new positions. If you're interested in working at the intersection of #technology and #security (with a little bit of #wargaming in the mix) here at the University of California, Berkeley, join us:
Berkeley Risk and Security Lab retweeted
Only had 12 hours on the ground in Boston but what an amazing way to spend it with the inaugural @MIT_SSP #Wargaming Lab Boot Camp led by the two Eric/ks @eriklg sharing how to think about rigid ruleset design and online fielding of wargames as experiments.
Berkeley Risk and Security Lab retweeted
Fantastic week in DC for @GCREAIM and @KissingerCenter @JHU_BIPR discussing #AI mil integration and all things #wargaming (not together ;). Always a pleasure to be with Frank and @HalBrands for the Kissinger Academy as well as @JaneVaynman!
Berkeley Risk and Security Lab retweeted
Last week was bittersweet as it was our very own @JDMenton's last day with us here @BerkeleyRisk as a postdoc as she transitions to her new role as a fellow at @CarnegieEndow. We can't wait to see her continued engagement in the #nuclear field.
Berkeley Risk and Security Lab retweeted
Great example of the value of really digging into cases that usually receive (at most) a cursory mention. Fascinating read from @NicoleGrajewski and @james_acton32 (and their threads about sourcing are possibly even more fun?)
In our new article for @ForeignPolicy, @james_acton32 and I wrote about the forgotten US-Soviet war scare of 1980.
Both the US and USSR believed the other had plans to invade Iran. Both misinterpreted each other's intentions, which fuelled the crisis. foreignpolicy.com/2024/06/09…
Berkeley Risk and Security Lab retweeted
Really appreciate the shoutout for some of our work at @BerkeleyRisk to characterize the use of mil-#AI in China via the leaked Zeus's Shield dataset led by @Ritwik_G
Full paper: arxiv.org/abs/2405.12167
The latest episode of the AI Policy Podcast is out - covering what happened at the Seoul AI Summit in Korea, the U.S.-China AI Safety Dialogues, and the discovery of a leaked Chinese military AI dataset.
Listen on all major podcast platforms:
csis.org/podcasts/ai-policy-…
Berkeley Risk and Security Lab retweeted
What data feeds the #military #AI models of the future? Really enjoyed seeing this paper come to life @BerkeleyRisk led by the inimitable @Ritwik_G.
China is building AI to target American naval vessels, but how well does it work? In our latest paper, “Open-Source Assessments of AI Capabilities,” we train a model on a leaked Chinese targeting dataset and provide a new methodology for net assessment. arxiv.org/abs/2405.12167
Berkeley Risk and Security Lab retweeted
China is building AI to target American naval vessels, but how well does it work? In our latest paper, “Open-Source Assessments of AI Capabilities,” we train a model on a leaked Chinese targeting dataset and provide a new methodology for net assessment. arxiv.org/abs/2405.12167
Berkeley Risk and Security Lab retweeted
🎧 Thank you to @areddie89 for joining me to discuss how AI will transform warfare. cfr.org/podcasts/impact-ai-w…
Berkeley Risk and Security Lab retweeted
It was a privilege to sit down with @JamesMLindsay to discuss all things #AI-military integration—particularly as efforts to produce governance mechanisms in this area start to bear fruit… (here’s looking at you Pol Dec and @UnderSecT)
🎧 Thank you to @areddie89 for joining me to discuss how AI will transform warfare. cfr.org/podcasts/impact-ai-w…
Berkeley Risk and Security Lab retweeted
On today's Lawfare Daily, @TylerMcBrien spoke with @areddie89 about wargaming as a tool to manage risk from war to climate—and beyond.
lawfaremedia.org/article/law…