Security teams now face a complex challenge as shadow AI and third-party models proliferate across organizations. An AI asset inventory serves as a central system cataloging every model, API, and workflow used by a company. Four main tool categories have emerged to address this: GRC platforms, MLOps tools, ITAM and CAASM systems, and connected AI security platforms. Each offers distinct capabilities for identifying and managing AI risks, though limitations exist in comprehensive coverage.
Disney has entered a multiyear partnership with Adobe to develop proprietary generative AI models using Adobe Firefly Foundry. The collaboration will train models on Disney intellectual property for use across parks, resorts, cruise lines, and StudioLAB. This follows earlier reports of potential job cuts at Disney, though the AI deal focuses on accelerating creative workflows rather than replacing human workers. The models will be tailored to Disney properties and designed to produce safe, production-ready outputs.
China's technology sector has experienced significant market volatility, with the SSE Star 50 index plunging 40% in less than four months. The selloff wiped out at least $480 billion in market value, with AI-related companies comprising 42% of China's total stock market valued at around $10 trillion. Beijing has mobilized state-backed funds and regulators to support the market, though the decline threatens plans to shift household savings from property into technology investments.
Scott Wiener, recently elected to Congress from San Francisco, is positioning himself as a key figure in federal AI policy. The newly minted legislator authored California's SB 1047 and SB 53, the country's first frontier AI legislation. Wiener cited a METR report suggesting Anthropic achieved recursive self-improvement as evidence that quick regulatory action is necessary. His proposed agenda includes establishing a strong federal AI regulator, implementing whistleblower protections, mandating testing requirements, and requiring kill switch capabilities.
The semiconductor sector showed strong momentum with the Philadelphia Semiconductor Index surging over 6% to reclaim 10,000 points on November 9. Morgan Stanley maintained a bullish outlook despite concerns about potential U.S. data center delays, noting that long-term AI investment demand remains intact. The firm forecasts total capex at $1.6 trillion in 2027 and $1.8 trillion in 2028, with about 20% of 2028 U.S. projects potentially facing delays. Morgan Stanley favors Nvidia, Broadcom, and TSMC for investment opportunities.
OpenAI's detection tools have proven effective in identifying AI-generated content, including a false graphic claiming Minnesota congresswoman Ilhan Omar won her district with 99 percent of the vote. The graphic, which spread after the 2026 midterm elections, actually showed Omar winning with roughly 76.5 percent against Republican John Nagel at about 16 percent. OpenAI's tool identified the image as AI-generated, highlighting ongoing challenges with AI-generated political content and the importance of detection capabilities.
UK finance employees are experiencing unexpected consequences from AI adoption, with some working increased hours rather than gaining time savings. A UK-based quant reported that technical staff weekly hours are rising despite AI implementation expectations. The findings suggest AI adoption in finance sectors may not always lead to the anticipated productivity improvements and could potentially increase workloads for some employees.
Geminus, a developer of physics-based AI technology for industrial operations, received a strategic investment from Bahrain's Bapco Energies through its venture arm BeVentures. The two companies signed a three-year partnership at ADIPEC 2026 to build a digital twin platform for Bapco Energies' energy infrastructure. The platform will combine operational data and AI-driven modeling to help engineers optimize decisions across gas and power systems, though specific financial terms of the investment remain undisclosed.
User history significantly impacts AI accuracy in news categorization tasks. Testing revealed that a user's own history improved accuracy by 2.3 percentage points over models with no history. However, when histories were matched on topic labels and text similarity, the performance gap narrowed to just 0.9 points. These findings suggest personalization gains may stem from topic overlap rather than true user-specific learning, offering insights for more effective AI customization approaches.
Key Takeaways
- Security teams need comprehensive AI asset inventories to manage shadow AI and third-party model risks across organizations.
- Disney partnered with Adobe to build proprietary AI models using Firefly Foundry trained on Disney intellectual property for parks and studio use.
- China's tech stock selloff erased $480 billion in market value, with AI companies representing 42% of a $10 trillion market.
- Scott Wiener's election to Congress positions him to shape federal AI policy with proposed regulations including kill switch mandates.
- Semiconductor stocks surged 6% as Morgan Stanley maintained bullish AI investment forecasts through 2028.
- OpenAI detection tools identified AI-generated political misinformation, including a false graphic about Ilhan Omar's election win.
- UK finance workers report increased hours despite AI adoption, challenging assumptions about productivity gains.
- Geminus secured investment from Bapco Energies to develop a digital twin platform for energy infrastructure optimization.
- User history improves AI accuracy in news categorization, but gains diminish when matched on topic similarity.
- AI governance requires addressing both shadow AI usage (57% of employees) and autonomous agent risks through behavioral monitoring.
Best tools to build and maintain an AI asset inventory
Managing AI tools is now a major challenge for security teams because shadow AI and third party models are hard to track. An AI asset inventory is a central system that catalogs every AI model, API, and workflow used by a company. The article describes four types of tools including GRC platforms, MLOps tools, ITAM and CAASM systems, and connected AI security platforms. Each type has strengths and limits in spotting and managing AI risks.
Best AI security platforms for 2026
The growth of generative AI and autonomous agents has changed how companies manage risk in 2026. Security teams must govern unapproved AI tools, secure non human identities, and control access to corporate data. The article says effective AI security platforms need shadow AI discovery, agent governance, data protection, and ITDR integration. These capabilities help companies enforce policies without slowing down innovation.
Top AI governance tools for shadow and agentic risks
Shadow AI and agentic AI create different governance problems that require strong visibility. Shadow AI is often used by employees without approval, and one report says 57 percent of employees did so in June 2026. Agentic AI can act on its own, so security teams need behavioral monitoring and clear audit trails. The article explains how platforms use existing tools, runtime controls, and risk registers to keep governance current.
Disney signs Adobe AI deal amid reported job cut plans
Disney signed a deal with Adobe to build proprietary AI models using Adobe Firefly Foundry. The models will be trained on Disney intellectual property and used mainly to speed up work in parks, resorts, cruise lines, and StudioLAB. Walt Disney Imagineering has already used custom image generation tools to cut project time. The deal follows a report that Disney is considering job cuts.
Disney and Adobe are building generative AI tools together
Disney and Adobe announced a multiyear partnership to create artist led generative AI models tailored to Disney properties. The deal gives Disney early access to Adobe technology for theme parks, cruise lines, ESPN content, and more. The models use unique character and environment data to produce safe, production ready outputs. The collaboration also includes exploring emerging technologies and responsibly developed AI.
China AI stock crash threatens Xi's tech funding plans
China's technology stock selloff has wiped out at least $480 billion in market value. The SSE Star 50 index plunged 40% in less than four months. AI-related companies make up 42% of China's total stock market value at around $10 trillion. Beijing is mobilizing state-backed funds and regulators to support the market. The decline threatens Xi Jinping's plan to shift household savings from property into technology investments.
Books to read instead of AI doom loops
This article recommends books for readers who want to think more clearly about AI's future. The author says today's AI technologies are genuinely dangerous to human agency and independent thinking. Doomsaying about AI has spread from researchers and CEOs into everyday media and conversations. The author argues that understanding technology's role in meaningful futures requires moving beyond fear-based narratives.
Scott Wiener ready to regulate AI in Congress
Scott Wiener won Nancy Pelosi's open San Francisco congressional seat and now could shape federal AI policy. He authored California's SB 1047 and SB 53, the first frontier AI legislation in the country. Wiener cited a METR report suggesting Anthropic achieved recursive self-improvement as evidence that Congress must act quickly. He wants a strong federal AI regulator, whistleblower protections, testing requirements, and kill switch mandates.
Semiconductor index jumps 6% as Morgan Stanley stays bullish on AI
The Philadelphia Semiconductor Index surged over 6% to reclaim 10,000 points on November 9. Morgan Stanley said the market overreacted to potential U.S. data center delays and that long-term AI investment demand remains intact. About 20% of 2028 U.S. projects may face delays, but total capex forecasts stay at $1.6 trillion in 2027 and $1.8 trillion in 2028. Morgan Stanley favors Nvidia, Broadcom, and TSMC for dip-buying.
AI graphic falsely inflates Ilhan Omar's election win
An AI-generated graphic falsely claimed Minnesota congresswoman Ilhan Omar won her district with 99 percent of the vote. Omar actually won re-election with roughly 76.5 percent of the vote against Republican John Nagel at about 16 percent. The graphic spread on social media after the 2026 midterm elections. OpenAI's detection tool identified the image as AI-generated. Omar won by a margin of more than 60 percentage points.
UK finance workers see AI increase working hours
AI was expected to save workers time, but some finance employees are working more hours because of it. A UK-based quant reported that technical staff are seeing their weekly hours go up. The findings show that AI adoption in finance may not always lead to time savings.
Geminus gets investment from Bapco Energies for AI partnership
Geminus, a developer of physics-based AI technology for industrial operations, received a strategic investment from Bahrain's Bapco Energies through its venture arm BeVentures. The two companies signed a three-year partnership at ADIPEC 2026 to build a digital twin platform for Bapco Energies' energy infrastructure. The platform will combine operational data and AI-driven modeling to help engineers optimize decisions across gas and power systems. Financial terms were not disclosed, and Bahraini teams will manage the technology.
Study examines how user history affects AI accuracy
A new study tested how replacing a user's history with another person's affects AI accuracy in news categorization. A user's own history improved accuracy by 2.3 percentage points over a model with no history. However, when histories were matched on topic labels and text similarity, the gap dropped to just 0.9 points. The findings suggest that personalization gains may come from topic overlap rather than true user-specific learning.
Sources
- Best Tools to Build & Maintain an AI Asset Inventory
- Best AI Security Platforms for 2026
- Top AI Governance Tools for Shadow & Agentic Risks
- Disney Signs Adobe AI Deal Amid Reported Job Cut Plans
- Disney and Adobe are building generative AI tools together
- China AI bust threatens Xi’s ’Wall Street East’ ambitions
- Essential Reading in an Era of AI Doomsaying
- Nancy Pelosi’s replacement is ready to regulate AI
- Philadelphia Semiconductor Index Jumps 6%, Reclaims 10,000: Morgan Stanley Says Long-Term AI Investment Logic Has Not Fundamentally Changed, Prefers Nvidia and Broadcom
- AI graphic amplifies US election conspiracies, inflates support for Ilhan Omar
- Who's reducing their weekly working hours the most in finance with AI: Revealed
- Geminus Secures Strategic Investment From Bapco Energies, Signs Three-Year AI Partnership
- Not Your Diary, Just Your Topic: Grading the Wrong-User Control
Comments
Please log in to post a comment.