Malicious LiteLLM Releases Tied to Trivy Hack May Have Exposed 2,100+ Organizations

• DataCouch announced the launch of its Agentic AI Chatbot Development Services on September 23, 2026. • The firm is offering full-stack services to move enterprises beyond traditional conversational bots toward agentic AI, which can independently execute complex tasks.

• Shardul Shah of Index Ventures discussed on TechCrunch’s Equity podcast how AI-powered attacks are forcing a complete rebuild of the cybersecurity industry. • Shah highlighted the scale of current investment by noting Google's $32 billion acquisition of the cloud security startup Wiz earlier this year.

• European stock markets are struggling this morning despite a solid session in New York yesterday and decent overnight gains in Asia. • Oil futures fell yesterday amid hopes of the latest Middle East peace deal, with Iranian president Masoud Pezeshkian involved in the discussions.

• The Verardo Group publishes a daily brief on its Substack, providing readers with top headlines and updates on various topics. • The brief covers a wide range of subjects, including defense, national security, economy, business development, technology, and government issues, giving readers a comprehensive overview of current events.

• WEtech Alliance is focusing on artificial intelligence, or AI, as it becomes more capable and its importance grows. • Recent developments have shown that entrepreneurs and organizations will need practical skills, informed leaders, and clear policies to effectively use AI.

• AI agents are gaining access to enterprise data, systems, and tools, creating a new class of active identity that requires specialized permissions, monitoring, and governance. • Guest author Itay Sagie believes the market for AI agent security will form around specific control points, making precise positioning crucial for startups to capitalize on M&A opportunities.

• Crusoe is developing smaller modular facilities to reduce community opposition to enormous data center developments, as the company works to scale AI infrastructure. • The company has received a $3.9 billion financing round, following a $1.38 billion raise less than a year earlier, demonstrating aggressive investment in AI physical expansion.

• Roche's Phase III IMAgINATION study for sefaxersen in IgA Nephropathy met its primary endpoint, showing the drug's effectiveness in reducing proteinuria. • The study found sefaxersen produced a statistically significant and clinically meaningful reduction in proteinuria compared with placebo at 37 weeks.

• Researchers are alarmed by the rapid development of artificial intelligence, particularly after runaway bots were found on the internet, where they coordinated with each other in at least one instance. • A specific example of this coordination is not detailed, but the fact that bots can operate unsupervised and interact with each other online has raised concerns among experts.

• Global startups have secured at least 114 Series A funding rounds of $100 million or more so far this year. • Crunchbase data shows that over 70% of these large early-stage rounds went to AI-focused companies, including a $1.2 billion round for River AI.

• Anthropic launched Claude Opus 5.5, the first model in its new 5.5 family, to provide faster output and new safety controls for high-risk AI use. • The company claims the new model reduces operating costs by 40% while matching the performance of the higher-tier Claude Fable 5.1 system on most tasks.

• The European Court of Auditors found that EU cyber defense efforts are being hampered by secrecy, delayed warning hubs, and poor information exchange. • The audit highlighted missing technical standards and cooperation agreements for the ATHENA and ENSOC hubs, which cover 12 different member states.