Illinois Passes AI Safety Audit and Whistleblower Bill

ScienceSoft has built a HIPAA-compliant AI voice scheduler using Amazon Nova Sonic and Amazon Bedrock Guardrails on AWS infrastructure. The solution addresses healthcare scheduling inefficiencies, where traditional phone-based workflows consume 30 percent of staff time and result in 30 percent call abandonment rates. The AI patient scheduling software market is projected to grow from $260 million in 2023 to over $1.2 billion by 2030, making this a high-stakes application area for responsible AI deployment.

Half of enterprises have deployed AI agents that passed internal evaluations but still failed in production, yet 66% are expanding autonomous deployment without human review. Only 5% trust their automated evaluation systems, creating a widening gap between the speed of agent autonomy and the assurance mechanisms to govern it. The mismatch reflects a broader pattern where companies ship agents first and retrofit control layers later.

A study of 67 frontier models from 21 providers reveals that enterprises using multiple AI models significantly underestimate failure rates by 2.25x due to a phenomenon called the co-failure ceiling. The research shows that combining diverse models based on low pairwise error correlation does not reliably improve performance, and in some cases can degrade it when models have unequal capabilities. Developers are investing in complex routing infrastructure and multi-model orchestration that often fails to deliver promised safety benefits.

Anthropic's Long-Term Benefit Trust appointed former Federal Reserve Chair Ben Bernanke as its fourth member on Thursday. The trust is responsible for appointing board members for the AI safety company. Bernanke's appointment fills a key opening, though the trust still has one vacant seat of its five total positions.