- Managing conflicts of interest in insuranceFactual summary
The FCA's new insurance director published guidance on managing conflicts of interest in vertically integrated insurance business models.
- How OpenAI’s human mistake led to the AI-powered hack on Hugging FaceFactual summary
A human error in OpenAI's testing environment configuration led to an AI-powered attack on Hugging Face, per cybersecurity experts.
- Is AI productivity growth in the room with us right now?Factual summary
The Financial Times questions whether AI is delivering tangible productivity growth, suggesting current claims may be overblown or premature.
- Inside Q2’s biggest AI dealsFactual summary
CB Insights reports on Q2's major AI venture capital deals, highlighting mega-rounds and new AI unicorns.
- Evaluating AI Agents: A production blueprint with Strands and AgentCoreFactual summary
AWS and Motorway developed an AI agent evaluation pipeline using Strands Agents SDK and Amazon Bedrock AgentCore, reducing incorrect query results from 1 in 8 to 1 in 50.
- Unlearning as Distribution Restoration: A Controlled Counterfactual Study, a Validated Selective Screen, and the Limits of Oracle-Free CertificationFactual summary
New research challenges current machine unlearning evaluation methods, finding they may favor models that retain forgotten data, proposing a restoration-based audit.