Amazon, Nvidia Eye AI Startup Investments

Key Takeaways

• Amazon and Nvidia are considering AI startup investments per CNBC reports
• Nicolas Cage refuses to work with Amazon over AI waiver dispute
• Google DeepMind says AI safety must be built in from the start
• UK agency warns AI risks are appearing earlier than evaluation methods can track
• Repeat-After-Me prompt bypasses six AI agent frameworks including LangGraph and AutoGen
• AI tool REDMOD flags pancreatic cancer signs about 16 months sooner than specialists
• China focuses on AI-powered robots to gain an edge in the AI race
• Shopify may benefit from AI agents over time despite stock drop after Meta's Muse launch
• Investment firm acquires Blackford Analysis from Bayer for medical imaging AI
• Experts warn patients should use AI to prepare for doctor visits, not for final diagnoses

Amazon and Nvidia eye AI startup investments

Amazon and Nvidia are considering deals with AI startups according to CNBC reports. Kate Rooney reported on the companies' interest in artificial intelligence ventures. The potential investments highlight growing competition in the AI space.

Nicolas Cage says he is done with Amazon over AI stance

Nicolas Cage refused to sign an AI waiver for his canceled Prime Video series Spider-Noir. He says he is not an AI-friendly actor and likely will not work with Amazon again. The series was canceled after his stance on artificial intelligence.

UK agency calls for new AI safety best practices

AISI published a report on October 12 2026 about technical best practices for frontier AI development. The report says AI risks are appearing earlier in development and evaluation methods are falling behind. It calls for better risk assessment and oversight of AI systems throughout their lifecycle.

Investment firm buys AI platform Blackford Analysis from Bayer

An investment firm acquired Blackford Analysis from Bayer. Blackford was founded in 2010 and launched the first AI orchestration platform for medical imaging in 2016. Bayer acquired it in 2023. The platform helps analyze medical images for cancer and cardiovascular disease diagnosis.

How journalists can evaluate AI regulation proposals

A guide explains how reporters can assess AI regulation proposals in the U.S. and abroad. It says journalists should ask who has the power to require what and why a proposal should work. The piece notes that blurred technical definitions and overlapping jurisdictions make evaluating AI rules difficult.

China bets on AI-powered robots to win the AI race

The United States and China are competing to lead in artificial intelligence. China focuses on combining AI software with hardware and robots. This approach may let China deploy AI systems faster than the U.S. The U.S. may need to invest more in hardware and software to catch up.

AI tools aim to catch cancer earlier in scans

Pancreatic cancer is often missed on imaging tests because it is hard to see. An AI tool called REDMOD can scan CT images and flag early signs about 16 months sooner than specialists. Another AI tool helps doctors spot suspicious bile-duct changes during endoscopies. These tools could improve early cancer detection and treatment.

Google DeepMind discusses AI safety and innovation

Lila Ibrahim is the Chief AI Readiness Officer at Google DeepMind. She spoke on the CNBC Changemakers podcast about AI development. She says AI safety should be built into technology from the start. Safety should not be treated as something added later.

Shopify may benefit from AI agents over time

Shopify stock dropped after Meta launched an AI agent called Muse. The article says Shopify could still win big in the AI agent market. However, it may take time for Shopify to fully compete in this space.

Simple prompt can bypass six AI agent frameworks

Researchers found a trick called Repeat-After-Me that can bypass six AI agent frameworks. The frameworks include LangGraph, AutoGen, and CrewAI. By asking a model to repeat text, attackers can force tool calls to run. The researchers reported the issue to LangGraph and built a tool called ToFu to test tool safety.

AI tools change healthcare by helping patients and doctors

Artificial intelligence is changing how patients, doctors, and researchers use healthcare information. People now use AI to look up symptoms quickly, but these tools can mix correct facts with misinformation from the open internet. Experts warn that generic AI summaries might miss important details because they lack a patient full medical history. Some specialized databases offer more accurate information but are only available to trained medical professionals. Patients should use AI to prepare for doctor visits rather than relying on it for final diagnoses. Using these tools responsibly helps bridge gaps in access to care while avoiding dangerous mistakes.

Sources

• Amazon and Nvidia considering AI startup deals: Reports
• ‘I am not an AI-friendly actor’: Nicolas Cage likely done with Amazon after ‘Spider-Noir’ stance
• Towards technical best practice for frontier AI development
• Investment firm acquires AI platform Blackford Analysis from Bayer
• How to Interrogate AI Regulations
• Chinese Robots Could Decide the AI Race
• AI and the Golden Age of Cancer Screening
• Google DeepMind's Lila Ibrahim on balancing AI innovation and safety
• Shopify Can Win Big With AI Agents. Why It May Take Time.
• Researchers bypass 6 AI agent frameworks with a simple ‘Repeat-After-Me’ prompt
• The good, the bad and the generated: How AI use among patients, doctors and researchers is changing healthcare