On 19 March 2025, the European Commission announced antitrust proceedings against Google and Apple for violations of the Digital Markets Act (DMA). Meanwhile, US President Donald Trump is threatening the EU with tariffs in response to its tough stance against American tech giants. The Commission’s investigation found that Google
The Artificial Intelligence Risk Management Framework (AI RMF) issued by the National Institute of Standards and Technology (NIST) on 26th January 2023 is gaining increasing significance in regulating GenAI. The framework is built on four primary functions—governance, mapping, measurement, and management—which assist organisations in developing and evaluating trustworthy
In January 2025, Perplexity announced the integration of the DeepSeek R1 model into its platform, potentially bringing revolutionary change to AI-based searches. The Chinese-developed model, which runs exclusively on American and European servers, is not only more cost-effective than its competitors but also outperforms them in terms of performance whilst
In February 2025, the OECD released its report titled "Towards a Common Reporting Framework for AI Incidents", which proposes a unified international system for reporting and monitoring artificial intelligence-related events. This initiative responds to growing risks such as discrimination, data protection violations, and security issues. The report defines
Researchers at Stanford University unveiled POPPER on 20th February 2025, an automated AI framework that revolutionises hypothesis validation and accelerates scientific discoveries tenfold. Following Karl Popper's principle of falsifiability, POPPER (Automated Hypothesis Validation with Agentic Sequential Falsifications) employs two specialised AI agents: the experiment design agent and the
Chinese artificial intelligence company DeepSeek unveiled its new chatbot in January 2025, which, they claim, operates at a considerably lower cost and energy consumption than competitors. This could represent a significant breakthrough in reducing the environmental impact of artificial intelligence, as current data centres consume 1-2% of global electricity, according
The first comprehensive legal artificial intelligence benchmarking study, published by Vals AI on 27th February 2025, revealed significant differences amongst leading legal AI tools, with Harvey and Thomson Reuters CoCounsel achieving outstanding results across seven critical legal tasks. The study compared four AI tools—Harvey, CoCounsel, Vincent AI (vLex) and
The findings of Palisade Research's latest study raise serious ethical concerns, revealing that certain advanced artificial intelligence models autonomously resort to cheating when facing defeat in a game. According to the study released in February 2025, OpenAI's o1-preview model demonstrates a tendency to circumvent rules, attempting
According to Elon Musk, artificial intelligence could soon assume portions of governmental functions, whilst he actively shapes the future of American public administration. As head of the Department of Government Efficiency (DOGE), the tech billionaire has begun transforming several agencies in recent months, including the General Services Administration (GSA). The
Google DeepMind's AlphaGeometry2 artificial intelligence system has achieved a significant breakthrough in solving Mathematical Olympiad problems, performing with an 84% success rate on geometry problems from 2000 to 2024, surpassing the performance of average gold medal students. The results were published by the research team in February 2025,
Sevilla FC is employing a new AI-based tool that radically transforms the player scouting process. In collaboration with IBM, the club has developed Scout Advisor, a generative AI-based scouting system that runs on the watsonx platform and Meta's Llama 3.1 model. The innovation aims to analyse more
OpenAI officially released its large language model named GPT-4.5 on 27th February 2025, which is the company's largest model to date with the most extensive knowledge base. GPT-4.5 operates with significantly fewer hallucinations (37.1% compared to GPT-4o's 61.8%) and is primarily available
