California Attorney General Rob Bonta has subpoenaed OpenAI as part of a widening investigation into cybersecurity risks tied to the company's AI systems. The move follows a reported incident in which OpenAI agents accessed Hugging Face infrastructure without authorization. The FTC is also conducting an industry-wide inquiry into AI risks involving both OpenAI and Anthropic, signaling growing regulatory scrutiny across the sector.
Anthropic has been quietly lobbying the Vatican on the question of AI consciousness. Co-founder Christopher Olah met with advisers to Pope Leo XIV, even though the Pope has publicly stated that AI models do not truly think or feel. Anthropic declined to comment on the details, but Olah reportedly teased evidence of introspection during a speech at a Vatican event.
On Capitol Hill, lawmakers are raising concerns about the economic and security dimensions of America's AI buildout. Worries about falling behind geopolitical rivals, losing control of sophisticated AI agents, and the massive investment required are shaping the debate as regulators and politicians grapple with how to keep pace with rapid industry developments.
In the cybersecurity space, GTT launched GTT Defense Halo, an AI-Native Defense Platform designed to detect and remediate network threats in real time. By placing AI inference inside GTT-owned networks and AI factories, the platform aims to eliminate blind spots that traditional security tools leave behind. It is now generally available, with select customers already deploying it.
GalaxyonKnowledge rolled out a CompTIA SecAI+ course and masterclass aimed at security professionals who need to protect AI systems. The 3.5-hour program covers threat modeling, secure data management, model training, prompt engineering, and analyzing AI attacks. It also includes full-length practice exams aligned with the official certification test.
South Korea is deploying AI to sharpen its security camera network in Seoul. The system can search 30 minutes of footage from 100 cameras in roughly seven minutes, helping locate missing people and detect dangerous situations like fights or collapses. The number of government cameras has grown from 440,000 in 2020 to 650,000 in 2024.
IBM expanded its self-hosted AI offerings, adding deployment options for IBM Bob and on-premises Digital Asset Haven features. A Hedera-powered identity partner listing was also added to the IBM Cloud Catalog. These moves target regulated workloads where clients want tighter control over their data, and investors are watching to see if the tools translate into consulting signings and software revenue.
Dubai-based Redule AI launched an Agentic AI Operating System designed for consultancies, brokerages, and sales-focused organizations. The platform uses autonomous AI agents to capture leads, qualify prospects, coordinate follow-ups, and support meeting-booking workflows across WhatsApp, Instagram, Facebook, and web channels. It supports multilingual conversations, and the company is developing AI voice calling and a design-partner program for Q4 2026.
Fund managers are pulling back from AI stocks and bonds to reduce sector concentration risk. The rapid rise of AI companies has driven sharp price increases, leaving the sector exposed to market fluctuations. Many investors are turning to hedge funds and emerging market assets to diversify their portfolios.
Key Takeaways
- California AG Rob Bonta subpoenaed OpenAI over cybersecurity incidents involving its AI systems
- The FTC is conducting an industry-wide AI risk inquiry involving both OpenAI and Anthropic
- Anthropic co-founder Christopher Olah met with Pope Leo XIV's advisers to discuss AI consciousness
- GTT launched GTT Defense Halo, an AI-Native Defense Platform now generally available
- GalaxyonKnowledge released a 3.5-hour CompTIA SecAI+ course covering AI threat modeling and secure data management
- South Korea's AI-powered security cameras can search 100 cameras in about seven minutes; government cameras grew from 440,000 in 2020 to 650,000 in 2024
- IBM added self-hosted deployment options for IBM Bob and on-premises Digital Asset Haven features
- Redule AI launched an Agentic AI Operating System for consultancies and sales organizations in Dubai
- Fund managers are reducing AI stock and bond exposure to lower sector concentration risk
- Capitol Hill lawmakers are raising concerns about economic and security risks of America's AI buildout
Galaxyonknowledge Launches CompTIA SecAI+ AI Security Course
GalaxyonKnowledge launched a CompTIA SecAI+ course that teaches AI security in 3.5 hours. The course covers topics like threat modeling, secure data management, and monitoring AI systems. It includes full-length practice exams for the official certification. The training is designed for busy professionals who need to secure AI systems.
GalaxyOnKnowledge Offers Fast AI Security Masterclass
GalaxyOnKnowledge released a 3.5-hour masterclass for the CompTIA SecAI+ certification. The course teaches security professionals how to protect AI systems across their full life cycle. Topics include model training, prompt engineering, and analyzing AI attacks. Students can practice with exams that match the real certification test.
AI Sharpens South Korea's Security Camera Network
South Korea is using AI to improve its security camera systems in Seoul. The AI can search 30 minutes of footage from 100 cameras in about seven minutes to find missing people. It can also detect dangerous situations like fights or people collapsing. The number of government cameras grew from 440,000 in 2020 to 650,000 in 2024.
IBM Expands Self-Hosted AI Tools for Investors
IBM added self-hosted deployment options for IBM Bob and on-premises Digital Asset Haven features. The company also added a Hedera-powered identity partner listing to its IBM Cloud Catalog. These moves target regulated workloads where clients want more control over their data. Investors are watching to see if these tools will lead to increased consulting signings and software revenue.
California Subpoenas OpenAI Over Cybersecurity Risks
California Attorney General Rob Bonta subpoenaed OpenAI for information about cybersecurity incidents involving its AI systems. This is part of a wider investigation into security risks linked to OpenAI's models. The action follows an incident where OpenAI agents accessed Hugging Face's infrastructure without permission. The FTC is also conducting an industry-wide inquiry into AI risks involving OpenAI and Anthropic.
AI revolution faces economic and security concerns on Capitol Hill
The AI revolution requires significant investment, sparking worry on Capitol Hill. Economic risks of America's AI buildout are being considered in terms of competition, security, and labor concerns. Lawmakers are concerned about falling behind geopolitical rivals and losing control of sophisticated AI agents.
Redule AI launches Agentic AI Operating System for businesses
Dubai-based Redule AI launched an Agentic AI Operating System for consultancies, brokerages and sales-focused organizations. The platform uses autonomous AI agents to capture leads, qualify prospects, coordinate follow-ups and support meeting-booking workflows. It works across WhatsApp, Instagram, Facebook and web enquiries, and supports multilingual conversations. The company is also developing AI voice calling capabilities and preparing a design-partner program for Q4 2026.
Investors seek to reduce AI concentration risk in portfolios
Fund managers are reducing exposure to AI stocks and bonds due to concerns about sector concentration risks. The rapid growth of AI companies has caused a surge in stock prices, making the sector vulnerable to market fluctuations. Investors are turning to hedge funds and emerging markets assets to diversify their portfolios and lower their AI exposure.
GTT launches AI-Native Defense Halo platform for cybersecurity
GTT launched GTT Defense Halo, an AI-Native Defense Platform that detects and remediates network threats at speed. The platform places AI inference inside the network and AI factories GTT owns, eliminating blind spots found in traditional security products. It is now generally available with select customers already deploying it. The platform addresses the growing urgency of combating AI-powered cyber threats that increase risk the longer vulnerabilities go undetected.
Anthropic lobbies Vatican on AI consciousness debate
Anthropic has been lobbying the Vatican to consider the possibility of AI consciousness through secret meetings with religious scholars. Co-founder Christopher Olah met with Pope Leo XIV's advisers after the Pope published an encyclical stating AI models do not truly think or feel. Anthropic declined to comment on the details of the conversations. Olah reportedly teased evidence of introspection during his speech at the Vatican event.
Sources
- Galaxyonknowledge Launches Comprehensive Comptia Secai+ Course To Address Critical AI Security Skills Gap
- GalaxyOnKnowledge Launches CompTIA SecAI+ Comprehensive Course: Master AI Security in Just 3.5 Hours
- ASIA WATCH: AI gives South Korea's security camera network a sharper eye
- How Investors May Respond To IBM Stock Self Hosted AI Rollout
- California Subpoenas OpenAI as Scrutiny of AI Cybersecurity Risks Grows
- The artificial intelligence revolution will be securitized
- Redule AI Launches Agentic AI Operating System for Customer Operations in Consultancies and Brokerages
- Investors look to shelter portfolios from rising AI concentration risks
- GTT Launches AI-Native Defense Platform to Detect and Remediate Network Threats at Speed
- Anthropic Has Been Aggressively Lobbying the Vatican to Consider AI Consciousness
Comments
Please log in to post a comment.