Brianna Rosen

"The most immediate threat is not the 'AI apocalypse' – where machines take over the world – but humans leveraging AI to establish new patterns of violence and domination over each other." My latest for @just_security on the AI revolution in drone warfare.justsecurity.org/89033/ai-and-t…

Governing AI requires international agreements, but cooperation can be risky if there’s no basis for trust. Our new report looks at how to verify compliance with AI agreements without sacrificing national security. This is neither impossible nor trivial.🧵 1/

“The gap between the aspirations of international ... law and the reality for people on the ground is greater and greater,” said @AdHaque110. “And that is a problem for the law, because law is supposed to achieve things in the real world.” theguardian.com/law/2025/jun/2…

We are delighted to announce that Jen Easterly, former Director of the US Cybersecurity and Infrastructure Security Agency (@CISAgov), has joined the School as a Visiting Fellow of Practice with the Oxford Programme for Cyber and Technology Policy.👇 ow.ly/uI1C50WfoJp

UPDATE: Iranian sources confirm that ceasefire with Israel expected to soon go into effect. According to one source, Israel will stop firing at 3.30am Jerusalem time, with Iran anticipated to follow suit. amwaj.media/en/media-monit…

Some really good analysis here, especially on the negative impacts to the NPT and IAEA inspections process, of US strike on Iran.

Trump struck Iran without any authorization of Congress. We need to immediately return to DC and vote on @RepThomasMassie and my War Powers Resolution to prevent America from being dragged into another endless Middle East war.

With rising conflict between Israel and Iran, Dr Brianna Rosen (@rosen_br) highlights the critical strategic challenges facing the US. 👇

Indefensible: Israel’s Unlawful Attack on Iran My latest @just_security

Israel's Deputy Attorney General repeats a longstanding confusion. 🧵 1. An 'armed conflict' triggers the application of protective rules of IHL to both sides of the conflict. It does not authorize the use of force by either side of the conflict (let alone by both sides).

Rather than repeating the term mindlessly, journalists should ask: "preemptive of what?" Absent an imminent & specific armed attack that cannot be avoided except with force, "preemptive" strikes are straightforwardly illegal. No attack meeting those criteria is even claimed. 1/4

Anthropic announced they've activated "Al Safety Level 3 Protections" for their latest model. What does this mean, and why does it matter? Let me share my perspective as OpenAl's former lead for dangerous capabilities testing. (Thread)

What does peace look like in the Russo-Ukraine War –and who gets to decide? Professor Sir Lawrence Freedman (@LawDavF) joined Professor @JaninaDill and Dr @rosen_br for a conversation on the diplomatic efforts to end the war. Catch up here 👇 youtube.com/watch?v=O4qN_R…

EVENT: Singapore’s cybersecurity playbook David Koh, Commissioner of Cybersecurity at @CSAsingapore, joins Prof Ciaran Martin as they explore Singapore’s strategy for responsible state behaviour in the digital realm. 📅 09 May, 3:30pm BST Register ⬇️ ow.ly/UaQW50VHFXc

We kicked off our new Calleva-Airey Neave Global Security Seminar Series with a superb keynote on "The evolution of global security threats since the Cold War" by Lord @WilliamJHague of Richmond, Chancellor of @UniofOxford, at the @BlavatnikSchool yesterday. We were so pleased…

bsg.ox.ac.uk/events/evoluti… Global Security is under threat now more than ever. To find out more about what’s happening, how we got here, and what to do moving forward, come along to the Calleva-Airey Neave Global Security Seminars. The first is on Monday.