OpenAI Co-Founder Brockman Defends $30 Billion Stake in Federal Trial
OpenAI co-founder and President Greg Brockman took the stand on May 4 in federal court in the Musk v. Altman trial. According to WIRED, Brockman revealed that he is one of OpenAI's largest individual stakeholders, with a stake worth approximately $30 billion. In his testimony, he said he had poured "blood, sweat, and tears" into the company, defending his position amid the controversy over its transition from a nonprofit to a for-profit entity.
The lawsuit was brought by Elon Musk, who alleges that CEO Sam Altman and others strayed from OpenAI's original nonprofit mission. It has become a landmark moment in which the battle over AI industry governance and direction is being argued directly in court.
White House Reportedly Weighs Pre-Release Review of AI Models
According to Forbes, the White House is exploring a framework for reviewing new AI models before they are released to the public. While details remain fluid at this stage, it is notable that mechanisms for proactively assessing AI safety and societal impact are now being concretely discussed.
As the pace of AI development accelerates, governments around the world are searching for the right regulatory approach. How the United States moves from discussion to implementation will be a key focus going forward.
Anthropic and FIS Partner to Build AI Agents for Combating Financial Crime
According to the Wall Street Journal, Anthropic and financial services firm FIS have announced a partnership to develop AI agents that help banks monitor and detect financial crime. The agents are designed to assist in identifying money laundering and fraudulent transactions.
AI adoption in finance is accelerating, with practical deployment advancing particularly in the security and compliance space. It is also a significant signal that Anthropic is getting serious about the enterprise AI agent market.
AWS Previews AgentCore Optimization, a Quality Management Tool for AI Agents
AWS has released a preview of "AgentCore Optimization," a new capability of Amazon Bedrock AgentCore. The tool automatically generates improvement suggestions from production trace data, and supports validating them through batch evaluations and A/B testing before deployment.
AI agents face a known challenge: once in production, their quality can quietly degrade as models are updated and user behavior shifts. Traditionally, this improvement cycle has relied on developers manually reviewing traces and rewriting prompts. AgentCore Optimization proposes an approach that shifts this cycle to data-driven automation.
Sierra Raises Approximately $1 Billion at a $15 Billion Valuation
According to Latent Space's AI newsletter, AI agent company Sierra has raised approximately $1 billion at a valuation of around $15 billion. The company announced that it surpassed $100 million in ARR in November 2025 and reached $150 million in February 2026, and it is now believed to have exceeded $200 million.
Sierra, which deploys AI agents for customer service, is drawing attention as a case study emblematic of the expanding enterprise AI market, with its rapid growth standing out in particular.
---