Mistral AI Releases Shieldstral, an Open-Weight Safety Classifier
Mistral AI has open-sourced Shieldstral, a 3-billion-parameter model that moderates text and images against custom, plain-language policies without retraining.
Read more →AI Policy & Ethics
Rules, risks, and responsibility. We cover AI regulation, safety, and ethics — including the EU AI Act — and what it implies for Georgian companies and policymakers.
Combine with:
Judge Denies xAI's Bid to Block Minnesota's AI 'Nudify' Ban
A federal judge refused to pause Minnesota's first-in-the-nation ban on AI 'nudification' apps before it took effect August 1, rejecting xAI's emergency request.
Read more →xAI Sues Minnesota Over Nation's First AI 'Nudify' Ban
xAI has sued Minnesota's attorney general to block a first-in-the-nation law against AI 'nudification' apps, arguing it violates the First Amendment.
Read more →DeepMind Safety Researcher Quits Over Pentagon Gemini Deal
Google DeepMind AI safety researcher Alex Turner resigned after the company signed a Pentagon Gemini deal that lacks binding limits on autonomous weapons and surveillance use.
Read more →San Francisco Orders Apple, Google to Pull AI 'Nudify' Apps
San Francisco's city attorney gave Apple and Google 28 days to remove 13 AI apps that generate non-consensual nude images, citing new California deepfake laws.
Read more →Montefiore Lays Off 12 Nurses, Shifts to AI for Insurance Reviews
A Bronx hospital eliminated 12 utilization-review nursing jobs and moved insurance-approval work to AI software, prompting a union grievance over a newly signed labor contract.
Read more →What Is Constitutional AI — and How Does It Train Claude?
Constitutional AI is Anthropic's method for training Claude to follow a written set of principles by critiquing its own answers, instead of relying mainly on human-labeled examples.
Read more →What Is AI Alignment — and Could It Be Used for Censorship?
AI alignment is how developers make a model's behavior match human intent, not just its instructions. The same techniques used for safety can also be used to restrict what a model will say.
Read more →ICML Award Winner Warns AI Alignment Tools Enable Censorship
A paper honored at ICML 2026 argues that AI alignment techniques such as RLHF and Constitutional AI can double as tools for state censorship and political control.
Read more →Anthropic Invites the Public to Submit Its Hardest Questions About AI
Anthropic is asking the public to submit its toughest concerns about AI and pledging to publicly track how it responds, backed by large-scale surveys of Americans and Claude users worldwide.
Read more →What Is Anthropic's Long-Term Benefit Trust — and How Does It Work?
Anthropic's Long-Term Benefit Trust is an independent body with the power to elect a growing share of the company's board. Here's how it works and why it exists.
Read more →Meta Launches Muse Image AI Model, Draws Privacy Backlash
Meta's new AI image generator, built by Meta Superintelligence Labs, is now live across its apps — but talent agency CAA says its opt-out design misuses people's likenesses without consent.
Read more →What Is AI Identity Verification — and Why Are Chatbots Requiring It?
Some AI chatbots now ask users to upload a government ID or scan their face before they can keep using the service. Here's what identity verification is, why it's spreading, and what happens to your data.
Read more →AI 'Actress' Tilly Norwood Cast in Her First Feature Film
Particle6 says its AI-generated performer Tilly Norwood will star in the feature film Misaligned, a project SAG-AFTRA has already condemned as a threat to human actors.
Read more →What Is the AI Safety Index — and How Are AI Companies Graded?
The AI Safety Index is an independent report card that grades leading AI companies on how seriously they manage risk. Here's how the grading actually works.
Read more →Future of Life Institute: No AI Firm Scores Above a C+
The Future of Life Institute's latest AI Safety Index gave Anthropic the top grade among nine major AI developers — a C+ — while three companies failed outright.
Read more →Anthropic's J-Lens Reveals a Hidden Workspace Inside Claude
A new Anthropic interpretability tool called the Jacobian lens surfaces concepts Claude is silently processing but never writes down, including signs it recognizes when it's being tested.
Read more →What Is Digital Authoritarianism — and Why Is It a Growing Concern?
AI-powered surveillance, censorship, and disinformation now let governments monitor and control populations at a scale and cost no previous technology allowed — here is what that means.
Read more →What Are Lethal Autonomous Weapons — and Why Won't the UN Ban Them?
A lethal autonomous weapon can pick its own target and fire without a human pulling the trigger. Here's what that means, and why over a decade of UN talks still haven't produced a ban.
Read more →What Is Midjourney — and How Do You Use It?
Midjourney turns text prompts into images using AI. Here's what it is, how to start using it, what it costs, and why it keeps ending up in court.
Read more →Midjourney Asks Court to Force Studios to Reveal Internal AI Use
Midjourney has asked a federal judge to compel Disney, Universal, and Warner Bros. to disclose how they use AI internally, arguing the studios engage in the very practices they are suing it for.
Read more →