Control Red Team - Research Engineer/Research Scientist

About the role

<div class="content-intro"><h2>About the AI Security Institute</h2> <p>The AI Security Institute is the world's largest and best-funded team dedicated to understanding advanced AI risks and translating that knowledge into action. We’re in the heart of the UK government with direct lines to No. 10 (the Prime Minister's office), and we work with frontier developers and governments globally.</p> <p>We’re here because governments are critical for advanced AI going well, and UK AISI is uniquely positioned to mobilise them. With our resources, unique agility and international influence, this is the best place to shape both AI development and government action.</p></div><h2><span class="TextRun SCXW212922402 BCX8" lang="EN-GB" data-contrast="none"><span class="NormalTextRun SCXW212922402 BCX8" data-ccp-parastyle="heading 2">The deadline for applying to this role is 30</span></span><span class="TextRun SCXW212922402 BCX8" lang="EN-GB" data-contrast="none"><span class="NormalTextRun Superscript CommentHighlightRest SCXW212922402 BCX8" data-fontsize="12" data-ccp-parastyle="heading 2">th</span></span><span class="TextRun SCXW212922402 BCX8" lang="EN-GB" data-contrast="none"><span class="NormalTextRun CommentHighlightRest SCXW212922402 BCX8" data-ccp-parastyle="heading 2">&nbsp;September 2026</span><span class="NormalTextRun CommentHighlightPipeRest SCXW212922402 BCX8" data-ccp-parastyle="heading 2">, end of day, anywhere on Earth.</span></span><span class="EOP Selected SCXW212922402 BCX8" data-ccp-props="{"134245418":true,"134245529":true,"335557856":16777215,"335559738":0,"335559739":0}">&nbsp;</span></h2> <h2>Team Description</h2> <p><span data-contrast="auto">Control measures — monitors, permission systems, sandboxing, resampling, escalation protocols — are designed to detect and prevent misaligned behaviour from advanced AI systems. Though the measures are already critical to safety, whether they would in fact catch a capable model attempting to cause harm is an empirical question that remains largely untested. </span><span data-ccp-props="{"335559739":160}">&nbsp;</span></p> <p><span data-contrast="auto">The Control Red Team at AISI is stress-testing control monitors (</span><a href="https://www.aisi.gov.uk/blog/how-our-new-control-red-team-is-stress-testing-frontier-monitors"><span data-contrast="none"><span data-ccp-charstyle="Hyperlink">including</span> <span data-ccp-charstyle="Hyperlink">from GDM and Anthropic</span></span></a><span data-contrast="auto">) – finding attacks that beat monitors and attempting to design experiments to measure real-world risk. We’re based within the Red Team, about a dozen people who have spent the last two years breaking developer’s alignment and misuse safeguards, and grew out of </span><a href="https://www.aisi.gov.uk/research?category_equal=%5B%22Control%22%5D"><span data-contrast="none"><span data-ccp-charstyle="Hyperlink">AISI’s</span> <span data-ccp-charstyle="Hyperlink">previ</span><span data-ccp-charstyle="Hyperlink">o</span><span data-ccp-charstyle="Hyperlink">us </span><span data-ccp-charstyle="Hyperlink">research</span><span data-ccp-charstyle="Hyperlink"> in</span><span data-ccp-charstyle="Hyperlink">to</span><span data-ccp-charstyle="Hyperlink"> contro</span><span data-ccp-charstyle="Hyperlink">l evaluations and safety cases.</span></span></a><span data-ccp-props="{"335559739":160}">&nbsp;</span></p> <p><span data-contrast="auto">We're opening roles on the Control Red Team, and we think it's an unusually good place to do this work. You'd join early, with real ownership over the team's direction; you'd have frontier model access, serious compute and strong infrastructure support from across AISI; and you'd get privileged insight into control measures across several frontier developers, working alongside some of the most experienced red teamers in the field.</span><span data-ccp-props="{"335559739":160}">&nbsp;</span></p> <p><span data-contrast="auto">Our current bet is to focus our eff

Skills