Blog

How to actually protect against digital sexualized violence

AlgorithmWatch has put forward recommendations on how to implement a ban of deepfakes in the AI Act as part of the AI Omnibus procedure. To effectively protect victims of digital sexualized violence, AI companies, platforms, and perpetrators must consistently be held accountable.

Close-up of a black computer keyboard showing three keys — X, C, and V — in the foreground. Below each key, the corresponding functions are labelled: "Cut", "Copy", and "Paste". The keyboard shows visible signs of use.

Copy, paste, govern: Microsoft ghostwrote EU policy that keeps data centers’ energy use secret

The EU Commission’s policy on data centers keeps information on the energy and water use of individual centers under wraps. Research by Corporate and Europe Observatory and AlgorithmWatch, published by Investigate Europe, reveals the Commission copied and pasted an amendment suggested by Microsoft and the lobby group Digital Europe. The aim: To prevent NGOs from obtaining information on energy-hungry data centers in the face of growing resistance.

Black-and-white photo of a person typing with a Chatbot on a laptop at a desk, showing only their hands and part of the keyboard; paperwork and books are stacked in the background.

Summary: Could AI Chatbots influence a Government’s Decisions?

What does it mean for democracy if our political leaders and government officials allow AI to shape their decisions?

Illustration depicting digital biometric identification and risks around it. Made for the entry "Personas expertas de todo el mundo alertan sobre riesgos de los sistemas de identificación biométrica digital" from the blog Contenido libre de R3D: Red en defensa de los derechos digitales.

The Seamless Surveillance Machine: Europe’s Biometric Border Vision

The EU aims to develop and deploy so-called “biometrics on the move” technologies in order to turn its borders into engines of seamless mass discrimination. Behind the promise of rendering border crossings instantaneous and border checks invisible, lies a vision rooted in opacity and unproven technological solutions. This could, in practice, usher in a regime of invisible but pervasive mass surveillance of people on the move and travelers alike.

Bird's eye view photo of a small hut and a concrete path through a lush green forest. However, the image is slightly distorted by digital artefacts.

The EU’s empty promises on sustainable border AI

EU-funded projects claim to make automated border surveillance environmentally sustainable, but the Commission admitted to AlgorithmWatch that it lacks rigorous definitions, standards and methods to translate these claims into measurable outcomes. As a result, a more ubiquitous “Green” border AI will most likely harm — rather than help — the planet.

The collage shows 4 archival images of women. In these of the images, the women are nude. There is also one portrait of a woman with yellow shapes and bounding boxes on her face.

Sexualized images on X: What we are doing to stop them and what we expect from the EU

X’s Grok chatbot is the focus of yet another scandal after generating pictures of real people in bikinis, without their consent, including children. But the problem of AI-generated sexual images without consent on X goes much further than Grok — and X blocked our research to address the problem. The EU Commission needs to step up their game to protect people from this kind of violence.

Within the frames are people bound to their office cubicles; beyond them, individuals work freely from diverse locations, connected through digital signals.

Algorithmic Accountability Reporting Fellowship

Fellows of AlgorithmWatch’s reporting program will research surveillance and discrimination

As we approach the end of the third year of the Algorithmic Accountability Reporting Fellowship, we are delighted to welcome a new cohort of eight journalists and researchers. This round will focus on two key clusters: digital surveillance and AI-driven intimate image abuse.

An array of colorful, fossil-like data imprints representing the static nature of AI models, laden with outdated contexts and biases.

Topic overview

Resource consumption of AI: The insatiable industry and its costs

Artificial intelligence does not simply fall from the sky. Its development, hardware production and everyday operation consume vast amounts of electricity, water and other resources. Yet tech companies remain reluctant to disclose the true costs to people and the environment. Here, we provide an overview of the issue.

Black and white photograph of a newspaper rack displaying German daily newspapers including Die Zeit and Süddeutsche Zeitung

Are Google AI Overviews killing media pluralism? AlgorithmWatch amongst first organizations to investigate that.

Google could destroy the web traffic which is the life-blood of organisations who produce reliable information. This is thanks to their new AI Overviews – a tool that hides links to real websites, in favour of their own summaries which have (for example) claimed Olaf Scholz is still the German Chancellor. AlgorithmWatch is making one of the very first data access requests under new EU rules to investigate these systemic risks to media freedom.

Image text: Maria Exner, Meredith Whittaker, Matthias Spielkamp.

Event

The Individual in the Machine – Meredith Whittaker on reclaiming Privacy in the Age of AI

On 25 September AlgorithmWatch invited to join an exclusive talk with Meredith Whittaker, President of the messenger platform, Signal. Meredith Whittaker, Maria Exner (Publix) und Matthias Spielkamp (AlgorithmWatch) discussed what we must do to bring technology in line with human needs - particularly in protecting the privacy of individuals from powerful platforms and operating system providers.

Automation on the Move

Border Surveillance on the Move to Enforce Restrictive Measures

In two recent Horizon Europe research projects, adaptable and mobile AI-based surveillance assemblages are developed to secure both the external and internal borders of the European Union. AlgorithmWatch looked into project material that revealed a lopsided fixation on defense.

Automation on the Move

The EU Spends Big on Border Tech — But Has No Idea What It Gets

The European Commission claims not to monitor if research findings of EU-funded projects are applied to the market after they have ended, AlgorithmWatch found. As no EU institution seems to be responsible for checking on border security investments, it is hard to tell if the millions spent actually lead to technological innovation.

Let’s Stop Nudification Apps Together!

Non-consensual nudity services are a horrifying use of AI. AlgorithmWatch is trying to use the EU’s Digital Services Act to limit the spread of non-consensual nudity services on social media and app stories. But platforms like X are standing in our way. What are our next steps and how you can help

A neural network comes out of the top of an ivory tower, above a crowd of people's heads (shown in green to symbolise grass roots). Some of them are reaching up to try and take some control and pull the net down to them. Watercolour illustration.

Open call to apply for AlgorithmWatch’s reporting fellowship on AI and power

For a fifth time, AlgorithmWatch is looking for new Algorithmic Accountability Reporting fellows. Apply now if you have research ideas concerning the relation between Artificial Intelligence and power and its consequences. The application deadline is 15 September 2025.

Pride With Pride! Stop Mass Surveillance at Pride, Stop Face Recognition Now

Report algorithmic discrimination!

When we apply for credit, apartments, or jobs online, companies increasingly use automated systems to process our data and make decisions that impact our daily lives. The problem: Such systems are not neutral and can reproduce inequalities and assumptions about people that already exist in society. What can we do to ensure that the use of non-transparent automated systems does not lead to people being disadvantaged? We need to make algorithmic discrimination visible, and you can help us with it!

What is algorithmic discrimination?

Discrimination and Artificial Intelligence (AI): Here's an overview of the topic.

Explainer: Predictive Policing

Algorithmic Policing: When Predicting Means Presuming Guilty

Algorithmic policing refers to practices with which it is allegedly possible to “predict” future crimes and detect future perpetrators by using algorithms and historical crime data. We explain why such practices are often discriminatory, do not hold up to what they promise, and lack a legal justification.

The Musk Effect: X’s impact on Germany’s election

AlgorithmWatch and the DFRLab have produced new research on X during German elections. We analyzed X posts by German politicians as well as prominent anti-far-right organizations, and found that the most viral posts are dominated by references to Elon Musk and his support for the AfD.

AI Action Summit in Paris – a missed opportunity?

Our Executive Directors, Angela Müller and Matthias Spielkamp, were in Paris last week representing us at the international AI Action Summit in Paris hosted by the French Government. So, what to make of the summit, the billion-dollar promises made at it, and the Big Tech party beats that never stop pounding? These are their main observations.

Simple network illustration.

Explainer: AI Energy Consumption

Fighting the Power Deficiency: The AI Energy Crisis

Is AI contributing to solving the climate crisis or to making it worse? Either way, the increase in AI applications goes hand in hand with the need for additional data centers, for which energy resources are currently lacking.

As of February 2025: Harmful AI applications prohibited in the EU

Bans under the EU AI Act become applicable now. Certain risky AI systems which have been already trialed or used in everyday life are from now on – at least partially – prohibited.

Mark Zuckerberg stepping towards a big X

Zuckerberg Makes Meta Worse to Please Trump

With his decision to gut moderation and fact-checking on Meta’s platforms, Instagram, Facebook and Threads, Mark Zuckerberg shows he cares more about the approval of Donald Trump than how his platforms can harm society.

Podcast False Positives Mockup with Mobile

False Positives: A Podcast on Financial Discrimination & De-banking

AlgorithmWatch and Agence France-Presse (AFP) released a podcast on automated discrimination in the financial sector, based in a six-month long investigation conducted in the framework of our Algorithmic Accountability Reporting fellowship.

A Year of Challenging Choices – 2024 in Review

2024 was a "super election" year and it marked the rise of generative Artificial Intelligence. With the adoption of the AI Act, it seemed poised to be the moment we finally gained control over automated systems. Yet, that certainty still feels out of reach.