Tootfinder

👍 D.C. Bar Committee Recommends that Jeffrey Clark, Who Tried to Help Trump Subvert the 2020 Presidential Election, Be Disbarred - Election Law Blog
https://electionlawblog.org/?p=151271

In the hands of someone like Trump,
deals are ways to evade, postpone or subvert the efficient work of markets.
Trump does not like markets,
precisely because they are impersonal and objective.
Their results – for corporations, entrepreneurs, investors and shareholders – are subject to clear measures of success and failure.
Because deals are personal, adversarial and incomplete,
they are perfect grist for Trump’s relentless publicity machine,
and a…

Concrete Security Bounds for Simulation-Based Proofs of Multi-Party Computation Protocols
Kristina Sojakova, Mihai Codescu, Joshua Gancher
https://arxiv.org/abs/2507.22705 https…

Concrete Security Bounds for Simulation-Based Proofs of Multi-Party Computation Protocols
The concrete security paradigm aims to give precise bounds on the probability that an adversary can subvert a cryptographic mechanism. This is in contrast to asymptotic security, where the probability of subversion may be eventually small, but large enough in practice to be insecure. Fully satisfactory concrete security bounds for Multi-Party Computation (MPC) protocols are difficult to attain, as they require reasoning about the running time of cryptographic adversaries and reductions. In this…

Möglicherweise sind die US Amerikaner gar nicht kollektiv verrückt geworden:
„She Won. They Didn't Just Change the Machines. They Rewired the Election.“
„The U.S. election wasn’t their endgame. It was their litmus test.“
https://substack.com/inbox/post/165658

She Won. They Didn't Just Change the Machines. They Rewired the Election.
How Leonard Leo's 2021 sale of an electronics firm enabled tech giants to subvert the 2024 election.

Thought Purity: Defense Paradigm For Chain-of-Thought Attack
Zihao Xue, Zhen Bi, Long Ma, Zhenlin Hu, Yan Wang, Zhenfang Liu, Qing Sheng, Jie Xiao, Jungang Lou
https://arxiv.org/abs/2507.12314

Thought Purity: Defense Paradigm For Chain-of-Thought Attack
While reinforcement learning-trained Large Reasoning Models (LRMs, e.g., Deepseek-R1) demonstrate advanced reasoning capabilities in the evolving Large Language Models (LLMs) domain, their susceptibility to security threats remains a critical vulnerability. This weakness is particularly evident in Chain-of-Thought (CoT) generation processes, where adversarial methods like backdoor prompt attacks can systematically subvert the model's core reasoning mechanisms. The emerging Chain-of-Thought Atta…

A part of The Accountant 2's character roster (the tech-wiz kids) is interesting in the relation between ‘modern surveillance’ and established genres. Surveillance cameras in the way? Hack or shut them down. Need an edge for the good guys during the showdown? Take over a bunch of US Air Force drones to: pinpoint the bad guys‘ location in a deserted area; conceal the actions of the heroes dealing with them. You either subvert it or remove it from the equation to make the human plot possib…

Imagine spending years as an intelligence officer, working diligently to subvert the Kremlin,
only to watch the US stand with Russia, Iran and North Korea,
as it did in February when it voted against a UN resolution condemning the invasion of Ukraine.
In that moment, Tim Weiner said: “You come to the realization, if you hadn’t already:
‘My God, the president of the United States has gone over to the other side. He has joined the authoritarian axis.’”

The Safety Gap Toolkit: Evaluating Hidden Dangers of Open-Source Models
Ann-Kathrin Dombrowski, Dillon Bowen, Adam Gleave, Chris Cundy
https://arxiv.org/abs/2507.11544

The Safety Gap Toolkit: Evaluating Hidden Dangers of Open-Source Models
Open-weight large language models (LLMs) unlock huge benefits in innovation, personalization, privacy, and democratization. However, their core advantage - modifiability - opens the door to systemic risks: bad actors can trivially subvert current safeguards, turning beneficial models into tools for harm. This leads to a 'safety gap': the difference in dangerous capabilities between a model with intact safeguards and one that has been stripped of those safeguards. We open-source a toolkit to est…