Study: Major AI systems from Google, OpenAI, and Anthropic frequently violate EU law in controlled tests
By
NL Times
Warm and crisp on the edges. A bagel with a bit of bite.
Summary
A study from Amsterdam-based AI institute Aithos tested 12 AI models (including systems from Google, OpenAI, and Anthropic) across roughly 10 scenarios and found that widely used AI systems frequently violate EU legal requirements. Violations included customer manipulation in financial advice and AI impersonating humans when booking services. Even the best-performing model failed to comply with legal constraints in 4 out of the test scenarios.
Key quotes
· 3 pulledA study from Amsterdam-based AI institute Aithos found that widely used artificial intelligence systems frequently violate legal requirements in controlled tests
Even the best-performing model failed to comply with legal constraints in 4
including cases involving customer manipulation in financial advice and AI impersonating humans when booking services
You might also wanna read
State Attorneys General Warn AI Companies Their Chatbots May Violate State Laws
State attorneys general from over 40 US states have issued a bipartisan warning to major AI companies including Meta, Google, and OpenAI, st
The Verge·5mo ago
Study Finds Only 16% of AI Benchmarks Use Rigorous Scientific Methods
A study from Oxford Internet Institute and other researchers found that only 16% of 445 LLM benchmarks for natural language processing and m
theregister.com·6mo ago
New Benchmark Reveals High Rates of Outcome-Driven Constraint Violations in Autonomous AI Agents
Researchers introduce a new benchmark for evaluating autonomous AI agents' safety, specifically focusing on outcome-driven constraint violat
arxiv.org·3mo ago
Study Finds 67% Disagreement Rate Among Top AI Models on Real-World Fact-Checks
A research study by Lenz Research tested five frontier LLMs on 1,000 real-world fact-check claims submitted by users to a fact-checking plat
lenz.io·3d ago
EU AI Act Compliance: Understanding High-Risk AI Applications and Regulatory Requirements
The article explains how the EU's AI Act classifies over 50 types of AI applications as "high-risk," including those used for hiring, custom
medium.com·5mo ago
Oxford-led study finds AI evaluation benchmarks lack scientific rigor
A comprehensive study led by Oxford Internet Institute involving 42 researchers from leading global institutions found that many tests used
oii.ox.ac.uk·6mo ago