The former CEO and COO of a health startup will spend years in jail after conducting a large-scale fraud scheme

Read More

Cisco has patched a zero-day vulnerability exploited by a Chinese APT group to compromise Nexus switches

Read More

If you’ve been reading my blog, you’ve noticed that I have written a lot about AI and democracy, mostly with my co-author Nathan Sanders. I am pleased to announce that we’re writing a book on the topic.

This isn’t a book about deep fakes, or misinformation. This is a book about what happens when AI writes laws, adjudicates disputes, audits bureaucratic actions, assists in political strategy, and advises citizens on what candidates and issues to support. It’s a book that tries to look into what an AI-assisted democratic system might look like, and then at how to best ensure that we make use of the good parts while avoiding the bad parts.

This is what I talked about in my RSA Conference speech last month, which you can both watch and read. (You can also read earlier attempts at this idea.)

The book will be published by MIT Press sometime in fall 2025, with an open-access digital version available a year after that. (It really can’t be published earlier. Nothing published this year will rise above the noise of the US presidential election, and anything published next spring will have to go to press without knowing the results of that election.)

Right now, the organization of the book is in six parts:

AI-Assisted Politicians
AI-Assisted Legislators
The AI-Assisted Administration
The AI-Assisted Legal System
AI-Assisted Citizens
Getting the Future We Want

It’s too early to share a more detailed table of contents, but I would like help thinking about titles. Below are my current list of brainstorming ideas: both titles and subtitles. Please mix and match, or suggest your own in the comments. No idea is too far afield, because anything can spark more ideas.

Titles:

AI and Democracy
Democracy with AI
Democracy after AI
Democratia ex Machina
Democracy ex Machina
E Pluribus, Machina
Democracy and the Machines
Democracy with Machines
Building Democracy with Machines
Democracy in the Loop
We the People + AI
Artificial Democracy
AI Enhanced Democracy
The State of AI
Citizen AI

Trusting the Bots
Trusting the Computer
Trusting the Machine

The End of the Beginning
Sharing Power
Better Run
Speed, Scale, Scope, and Sophistication
The New Model of Governance
Model Citizen
Artificial Individualism

Subtitles:

How AI Upsets the Power Balances of Democracy
Twenty (or So) Ways AI will Change Democracy
Reimagining Democracy for the Age of AI
Who Wins and Loses
How Democracy Thrives in an AI-Enhanced World
Ensuring that AI Enhances Democracy and Doesn’t Destroy It
How AI Will Change Politics, Legislating, Bureaucracy, Courtrooms, and Citizens
AI’s Transformation of Government, Citizenship, and Everything In-Between
Remaking Democracy, from Voting to Legislating to Waiting in Line
How to Make Democracy Work for People in an AI Future
How AI Will Totally Reshape Democracies and Democratic Institutions
Who Wins and Loses when AI Governs
How to Win and Not Lose With AI as a Partner
AI’s Transformation of Democracy, for Better and for Worse
How AI Can Improve Society and Not Destroy It
How AI Can Improve Society and Not Subvert It
Of the People, for the People, with a Whole lot of AI
How AI Will Reshape Democracy
How the AI Revolution Will Reshape Democracy

Combinations:

Imagining a Thriving Democracy in the Age of AI: How Technology Enhances Democratic Ideals and Nurtures a Society that Serves its People

Making Model Citizens: How to Put AI to Use to Help Democracy
Modeling Citizenship: Who Wins and Who Loses when AI Transforms Democracy
A Model for Government: Democracy with AI, and How to Make it Work for Us

AI of, By, and for the People: How Artificial Intelligence will reshape Democracy
The (AI) Political Revolution: Speed, Scale, Scope, Sophistication, and our Democracy
Speed, Scale, Scope, Sophistication: The AI Democratic Revolution
The Artificial Political Revolution: X Ways AI will Change Democracy…Forever

Read More

The EU Commission said Meta’s pay or consent model means users cannot freely consent to their personal data being collected for advertising purposes

Read More

A newly discovered RCE vulnerability, which can lead to full system compromise, has put over 14 million OpenSSH server instances are potentially at risk, according to Qualys

Read More

A new paper, “Polynomial Time Cryptanalytic Extraction of Neural Network Models,” by Adi Shamir and others, uses ideas from differential cryptanalysis to extract the weights inside a neural network using specific queries and their results. This is much more theoretical than practical, but it’s a really interesting result.

Abstract:

Billions of dollars and countless GPU hours are currently spent on training Deep Neural Networks (DNNs) for a variety of tasks. Thus, it is essential to determine the difficulty of extracting all the parameters of such neural networks when given access to their black-box implementations. Many versions of this problem have been studied over the last 30 years, and the best current attack on ReLU-based deep neural networks was presented at Crypto’20 by Carlini, Jagielski, and Mironov. It resembles a differential chosen plaintext attack on a cryptosystem, which has a secret key embedded in its black-box implementation and requires a polynomial number of queries but an exponential amount of time (as a function of the number of neurons). In this paper, we improve this attack by developing several new techniques that enable us to extract with arbitrarily high precision all the real-valued parameters of a ReLU-based DNN using a polynomial number of queries and a polynomial amount of time. We demonstrate its practical efficiency by applying it to a full-sized neural network for classifying the CIFAR10 dataset, which has 3072 inputs, 8 hidden layers with 256 neurons each, and about 1.2 million neuronal parameters. An attack following the approach by Carlini et al. requires an exhaustive search over 2^256 possibilities. Our attack replaces this with our new techniques, which require only 30 minutes on a 256-core computer.

Read More

Evil twin Wi-Fi access points mimicked legitimate networks to capture personal data from unsuspecting victims who mistakenly connected to them

Read More