Legal
AI Learns Legal Language to Predict Court Outcomes
A new study shows that teaching AI the structure of legal documents improves its ability to predict court decisions, offering a tool for faster legal analysis without costly retraining.
AI Learns When to Answer Legal Questions and When to Stay Silent
A new method combines smarter document retrieval with AI alignment to reduce errors in legal AI systems, enabling more reliable and safe automated legal analysis.
EU Privacy Law Demands Clear Consent for Online Tracking
A new legal analysis reveals that companies must obtain explicit user permission for behavioral advertising, challenging common opt-out practices and reshaping digital marketing rules.
EU Privacy Rules Need Urgent Update for Digital Age
A new analysis reveals how Europe's outdated electronic privacy laws fail to protect users in today's diverse online world, leaving gaps that could undermine fundamental rights.
AI Can Now Verify Long Documents in Real Time
A new method allows AI systems to check the accuracy of generated answers against entire documents, not just snippets, making enterprise tools more reliable without slowing them down.
Privacy as a Fairness Tool: How Local Differential Privacy Can Reduce Bias in AI
In the high-stakes world of algorithmic decision-making, where machine learning models determine everything from loan approvals to hiring, two critical concerns often seem at odds: fairness and privac…
AI Can Now Forget What It Learned
New machine learning methods enable AI systems to delete personal data on demand while maintaining performance—crucial for GDPR compliance and user privacy protection.
AI Outperforms Humans in Legal Document Labeling
New AI models accurately classify thousands of legal document labels, setting records in speed and precision for European Union law databases.
AI Learns When to Stop Accepting Risk
New research reveals optimal decision policies for AI systems facing selective data gaps, with implications for lending, hiring, and criminal justice applications.
AI's Seven Levels of Legal Reasoning
A new framework helps clarify AI's role in law, from simple automation to superhuman reasoning, offering a clear path to assess progress and prevent misleading claims.
AI Cuts Legal Data Noise by 73%
AI slashes legal document confusion by 73% - cutting through dense paperwork to expose criminal networks with unprecedented clarity
AI Maps Human Smuggling Networks from Legal Texts
AI now maps human smuggling networks from legal documents with 45% fewer errors, creating clearer visualizations for law enforcement. This tool helps uncover hidden criminal patterns buried in complex texts.
AI Personhood Is a Tool, Not a Truth
AI personhood isn't about consciousness - it's a practical tool for accountability and problem-solving. This revolutionary framework helps societies integrate AI responsibly without endless philosophical debates.
AI Now Checks Legal Rules More Accurately
AI now spots legal violations with unprecedented accuracy, saving businesses from costly compliance mistakes and regulatory fines.
AI Accountability: A Modern Magna Carta Challenge
As artificial intelligence systems grow more powerful and pervasive, their ability to influence financial decisions, control physical devices, and shape organizational processes raises urgent question…
AI Generates Certified Code for Critical Systems
In safety-critical industries like automotive and aviation, software must meet strict regulatory standards, but current AI tools often produce code that violates these rules, requiring costly manual f…
Self-Driving Cars Learn Traffic Laws Like Humans
Self-driving cars now understand traffic laws like humans, making complex legal decisions in real-time. This breakthrough could finally make autonomous vehicles safe enough for everyday roads.
AI Fails Real-World Long-Text Challenges
Large language models (LLMs) are increasingly used in applications requiring analysis of lengthy documents, from legal cases to financial reports, yet their ability to handle real-world long-context t…
AI Spots Unfair Contract Terms Efficiently
Online service agreements, often dense and unread, can hide unfair clauses that strip users of rights, making automated detection vital for consumer protection and regulatory oversight. A new study ev…
Public Interest Lawsuits Challenge Automated Injustice
Australian researchers reveal how strategic litigation holds automated decision-making accountable, forcing governments and companies to address algorithmic harms without waiting for new regulations.
AI Models Can Now Tune Their Own Transparency
Researchers develop a 'locality dial' that allows AI systems to adjust their interpretability levels without retraining, offering new possibilities for regulated industries where both accuracy and transparency are essential.
AI Watermarks Fail EU Standards
A new study reveals no current method for marking AI-generated text meets the EU's strict requirements for reliability and robustness, raising concerns over trust and regulation.