Image via TechCrunch
Monday, November 3, 2025
Sam Altman: Stop Asking About Our Money
Sam Altman is done answering questions about OpenAI's finances (yikes), while the company just launched Aardvark, a GPT-5 powered security agent that automatically hunts vulnerabilities. Meanwhile, Perplexity is having quite the week signing Getty Images despite ongoing copyright lawsuits, and Meta dropped their "Rule of Two" framework to protect AI agents from prompt injection attacks. Oh, and Google AI Studio now offers free logging and dataset tools because apparently debugging AI apps needed to get easier. Would you trust an AI agent securing itself?
Image via TechCrunch
Top Stories
TechCrunch
OpenAI's Sam Altman dismissed financial concerns over the company's massive infrastructure spending, claiming revenue well exceeds $13 billion and projecting continued steep growth across multiple business lines toward a potential $100B revenue target by 2027.
OpenAI
OpenAI's Aardvark uses AI agents to autonomously discover and fix code vulnerabilities at scale, representing a shift toward AI-powered continuous security that addresses the growing CVE crisis without slowing development.
Perplexity secured a Getty Images licensing deal to legitimize its image use and address copyright accusations, signaling a broader industry shift toward formal content partnerships rather than aggressive scraping practices.
Meta AI
Meta proposes the 'Agents Rule of Two' framework to secure AI agents against prompt injection attacks by limiting agent capabilities to two of three high-risk properties per session, addressing a critical vulnerability as autonomous agents gain access to sensitive systems and data.
Google Blog
Google AI Studio's new logging and datasets tools help developers monitor AI application quality and debug issues in real-time, enabling faster iteration and more reliable production deployments.
Keep Reading
Industry Voices
Enjoyed this issue?
Get daily AI intel delivered to your inbox. No fluff, just the stories that matter.