Digitalgipfel: Die letzte Warnung
Auf dem Digitalgipfel mahnt die KI-Taskforce, Deutschland verliere ohne rasches Handeln den Anschluss. Der Gipfel allein wird daran leider nichts ändern.
AI regulation
The last 7 days: stories about the topic itself first, then its parts, then where it is backgroundFollow it to see its stories first on the front page.
Postman silently auto-installs AI plugins and attempts to share all API data
Postman is an API platform for building and using APIs. Postman simplifies each step of the API lifecycle and streamlines collaboration so you can create better APIs—faster. - [SEVERE] Desktop auto...
One in 15 adults has experienced AI-generated image-based abuse, study finds
Around one in 15 adults has experienced AI-generated image-based abuse, according to a new study. The general-population survey of more than 7,200 adults in the United States, the United Kingdom and Australia found 7% had experienced at least one…
Forget retraining - AI can now tweak another AI's weights
The outputs of AI models are the result of a complex interplay between neurons, weights, training, and inference, and are therefore difficult to follow. But in some ways they're starting to look more like open books. Recent work by Anthropic safety…
Künstliche Intelligenz: Der KI-Elefant im Porzellanladen der Demokratie
Die großen Techplattformen beherrschen die Öffentlichkeit, und sie verfolgen dabei eine antidemokratische Agenda. Das kann so nicht weitergehen.
Google says there’s no way to use Search without using artificial intelligence
In announcement geared to schools, tech giant showed how entwined AI is with the world’s most popular website Google just acknowledged that you can’t use Google Search without using artificial intelligence. The search giant announced a slate of new…
Open models can do 90 percent of agentic work at a twentieth of the cost of a frontier LLM: Gartner
For-profit model-makers, many of which are spending billions on AI infrastructure, will be chasing revenue from just ten percent of steps involved in agentic AI workflows, according to analyst firm Gartner. “Nine in 10 agent workflow steps are…
AI ‘godfather’ Geoffrey Hinton: it might turn out we were better off without AI
Nobel winner warns the tech could be near to triggering major incident such as power or banking shutdown One of the “godfathers of AI”, Geoffrey Hinton, has said it could turn out that humans would have been better off never developing the…
Dollar falls while gold and US Treasury bonds rally as markets react to Democrats’ midterm election victory – business live
Democrats’ capture of Congress raise expectations of tougher regulation of AI sector; US stock futures fall but later pare losses Chris Beauchamp, chief market analyst at the investing and trading platform IG, says: Overnight markets had pointed…
AI is changing what's in store for new hires and managers alike
AI is screening résumés, prompting some managers to hire fewer entry-level workers and leading others to expect beginners to tackle more advanced work, Northeastern University's AI & Society Survey found. The report is the latest in a series that…
Google responds to political backlash over removing search from schools that ban AI
After elected officials voiced outrage over Google turning off search at schools that ban AI, the company said search can be turned back on with AI Overviews.
„Ultimatum ostil”. Google pune școlile în fața unei alegeri controversate: acceptă inteligența artificială sau renunță la motorul de căutare
Google tocmai a recunoscut că nu se poate folosi motorul de căutare Google fără a recurge la inteligența artificială, arată The Guardian. Marți, gigantul din domeniul căutării a anunțat o serie de noi produse bazate pe IA destinate sălilor de clasă…
Grammar-Constrained Decoding Can Jailbreak LLMs into Generating Malicious Code
arXiv:2606.11817v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly used for code generation, raising concerns that they may be misused to produce malicious code. Meanwhile, Grammar-Constrained Decoding (GCD) has…
Google will let schools skip AI tools, but they have to give up search, too
While Google is trying to move beyond Chromebooks, the ubiquity of these devices has made the company's education products huge in schools. Now, Google says it will let educators choose whether to allow its AI tools in the classroom. However…
Tell this AI what you want a protein to do—and it designs one
In a short space of time, artificial intelligence has become a whiz at designing proteins, molecular machines that perform various functions such as speeding up reactions, producing fluorescent signals and breaking down plastic. But getting AI to…
Is AI a Bubble? Wall Street Titans Say Not Even Close
Finance heavyweights including Goldman CEO David Solomon, Franklin Templeton's Jenny Johnson, Morgan Stanley's Ted Pick broadly dismissed fears that the AI boom has entered bubble territory, while Pimco chief Emmanuel Roman noted the sector's debt…
Semnale de alarmă cu privire la folosirea excesivă a AI-ul fără o direcție clară în imaginile de promovare
Agențiile de marketing online avertizează că grăbirea producției de bannere, postări, imagini pentru promovare și meniuri în restaurante generate cu AI erodează încrederea publicului în afacerile respective,...
Black Forest Lab on AI’s Next Phase
Andi Blattman, Co-Founder, Black Forest Lab discusses where lasting AI value will be created, who is positioned to capture it and the enterprise, infrastructure, safety and policy forces shaping the next phase with Bloomberg’s Shirin Ghaffary at…
The Pain Axis: LLMs Represent Self-Directed Harm and Act on It
arXiv:2609.16247v3 Announce Type: replace Abstract: LLMs sometimes behave in ways resembling human emotional responses, and recent work identified internal representations that may explain these behaviors. We ask whether LLMs represent pain…
Voluntary Servitude: Weaponized Safety Warning Mechanisms for Controlling AI Agents
arXiv:2611.04393v1 Announce Type: cross Abstract: LLM agents increasingly act on information retrieved from public sources they treat as credible, which turns a manipulated recommendation into a transaction, a dependency, or a decision with…
Step-Zero Attack: Agent Safety Before Observability
arXiv:2611.01953v1 Announce Type: cross Abstract: Existing attacks and defenses for coding agents act once content enters the model context: injected text persuades the model, and guards inspect what the model reads. Yet to judge whether processing…
EEFA: A Framework for the Ethical Evaluation of AI Agents
arXiv:2611.02561v1 Announce Type: new Abstract: The rapid deployment of autonomous AI agents across consequential domains including healthcare, legal services, nancial management, and enterprise workows has created an urgent governance challenge…
Here's Why You Shouldn't Use AI to Vote
From hallucinations to partisan bias, artificial intelligence chatbots like ChatGPT and Claude are not helpful when it comes to Election Day.
AI policy wonks declare 'sovereign agents' a threat to civilization
The Center for a New American Security posits that the next big threat from our future AI overlords involves "sovereign agents" acting on their own and regularly improving themselves. These are AI agents that operate outside of human control and…
Anthropic is working on a system to prove AI models’ identity
Called FlashProve, the system would let anyone check that an AI response really came from the model it claims to be from.
EuGH-Gutachten: KI-Chatbots können Urheberrecht verletzen
Presseverlage und Urheber wehren sich gegen die Nutzung ihrer Inhalte durch generative KI-Modelle. Im ersten EuGH-Verfahren zu dem Thema gibt es nun Hinweise.
Temperature, Not Knowledge: What Recursive Self-Training and Its Remedies Do to a Language Model's Confidence
arXiv:2611.01059v1 Announce Type: new Abstract: Recursive self-training can collapse language models, and many proposed remedies target overconfidence, yet their effects on confidence are rarely measured. We study a five-round self-training loop on…
Netflix will show you themed recommendations through a new ‘AI Discover’ tab
Netflix's TV app is getting a new "AI Discover" tab where the company will show experimental, AI-powered features to help you find new shows and movies to watch. The first of these experiments surfaces content based around themes, like "Something…
Bursă. Deranj tot mai mare pe bursa americană cu firmele de AI: Intel are minus 53% faţă de maxime, Google 31% şi AMD 22%. Toată lumea se întreabă dacă este posibilă o situaţie precum criza dot-com de la începutul anilor 2000. Ce semnale transmit scăderile?
Nota de plată pentru febra inteligenţei artificiale începe să ajungă pe Wall Street, unde investitorii nu mai privesc cu aceeaşi îngăduinţă miliardele de dolari pompate în centre de date, cipuri şi infrastructură.
Apple is all in on AI — its own and everyone else’s
Apple's latest earnings show how it's continuing to capitalize on the AI boom without busting its budget on the infrastructure spending consuming its competitors. Overall, Apple posted its best-ever fiscal year for revenue and earnings per share…
ASTĂZI Maratonul AI, Cloud & Resilience. AI în business: oportunități, infrastructură și riscuri, Private, public & hybrid cloud: control și flexibilitate, Cybersecurity și protecția datelor în era AI, Edge computing: performanță și proximitatea datelor
AI în business: oportunități, infrastructură și riscuri, Private, public & hybrid cloud: control și flexibilitate, Cybersecurity și protecția datelor în era AI, Reziliență digitală: multi-DC și geo-redundanță, Backup imuabil, recuperare și…
Revolut moves to self-hosted GLM
BREAKING: Revolut moved ~2,000 engineers onto self-hosted GLM for coding and pushed Claude Code and Codex to the exception list. Per David Abitbol, who leads AI ops there: 56 Nvidia B200s, a blended 30 cents per million tokens, roughly a tenth of…
US Stocks Waver as Election, AI Bubble Risks Weigh on Outlook
US stock futures were little changed as Americans headed to the polls on Election Day, with control of both chambers of Congress at stake and investors weighing mounting risks to the artificial intelligence trade.
STLT: A Novel System for Energy-Aware Distributed Inference in the Edge-Cloud-Space Continuum
arXiv:2611.01233v1 Announce Type: cross Abstract: Low Earth Orbit satellites are pivotal for Earth observation tasks such as wildfire detection, capturing imagery for analysis by Deep Neural Networks (DNNs). On-board analysis reduces downlink…
‘A serious failure’: CPS prosecutor cites imaginary case to high court after AI search
Exclusive: Nonexistent precedent cited in alleged child rape case adds to dozens of AI-related errors plaguing court proceedings A senior crown prosecutor acting in a high court case cited an imaginary legal precedent in an error that has sparked…
Scammers using AI-generated images of missing pets to demand reward money
Owners says threats to kill their cats unless they sent money left them angry and terrified.
VIDEO Acum suntem ajutați doar de AI. Marian Sîmpetru, Managing Partner eSolutions: Echipele care nu folosesc AI ajung să fie înlocuite de echipe care folosesc AI. Cel mai important client nou este din Germania - Profit.ro Live TV
Adopția inteligenței artificiale trece de la experimente la integrarea în procesele de business, iar companiile care întârzie riscă să piardă competitivitate, spune Marian Sîmpetru, Managing Partner eSolutions. Compania raportează o creștere a…
Black-Box Adversarial Example Attacks on Machine Learning based Election Systems
arXiv:2611.02166v1 Announce Type: cross Abstract: Machine learning models have shown exceptional accuracy for ballot classification in election systems. Recent research has demonstrated that these ballot classification models are vulnerable to…
Anunț surpriză al Revolut. A trecut la un model AI chinezesc renunțând la Anthropic și OpenAI. Ce schimbări s-au întâmplat deja
Compania de fintech Revolut folosește acum modelul cu ponderi deschise (open-weight) GLM-4.5 de la Zhipu AI, care rulează pe 56 de GPU-uri Nvidia B200, reducând costurile de inferență la aproximativ 0,30 USD per milion de tokeni, anunță un…
Alegeri SUA: Prima rundă de vot de importanță globală care se derulează în epoca Inteligenței Artificiale. Cum s-au pregătit Open AI, Anthropic și Google
Întors recent dintr-o vizită în China, editorialistul american Thomas L. Friedman a mărturisit că, pentru prima oară, a văzut un afiș pe aeroportul din Beijing. Pe el scria că posesorii de drone trebuie să le prezinte la control. Mi-am dat seama, a…
Google va întrerupe accesul la căutări pentru elevii din școlile care resping inteligența artificială
Compania Google, s-a confruntat cu opoziția profesorilor, a experților în dezvoltarea copilului și a părinților cu privire la utilizarea tehnologiei sale în școli, reacțiile negative crescând de când a anunțat în august că a activat chatbot-ul său…
I use AI (Oct 2026)
I document how I use AI, one year after my first writeup. Everything's completely different.
OpenAI hires former US Army secretary as military seeks to expand AI use
Dan Driscoll to join the start-up as senior counsellor for national security modernisation
Rusia dezvoltă sisteme capabile să bruieze Starlink pe suprafețe extinse. Cum funcționează noua amenințare
Forțele ruse își sporesc numărul sistemelor de război electronic (REB) concepute pentru a perturba comunicațiile prin satelit. Obiectivul este blocarea terminalelor Starlink de-a lungul frontului și în teritoriile ocupate de Rusia.
How AI giants could raise the cost of borrowing in Europe
France, Germany and Italy face a volatile bond market as war inflation bites. With US tech giants now issuing billions in AI bonds, they are competing with cash-strapped European governments.