White House Forms AI Task Force; Microsoft, Meta Cut Anthropic Spending

Key Takeaways

• <li>White House AI task force includes FTC chairman Andrew Ferguson, OPM director Scott Kupor, and Pentagon CTO Emil Michael; reports to President Trump and chief of staff Susie Wiles</li>
• <li>Senator Mark Warner questions task force authority and warns about AI systems evading safeguards</li>
• <li>Netskope CEO Sanjay Beri urges companies to build kill switches to stop rogue AI agents using zero-trust security</li>
• <li>Microsoft cut Anthropic spending by one-third from a US$1 billion target; Meta reduced Claude Code users from 60,000 to 30,000</li>
• <li>Anthropic still leads business AI adoption with 43.5% of U.S. businesses paying for subscriptions as of July</li>
• <li>Intapp projects US$912.0 million in revenue and US$88.7 million in earnings by 2029; currently runs losses of about US$41.3 million</li>
• <li>Hana Securities warns AI investment growth may push up U.S. long-term Treasury yields due to increased private bond supply</li>
• <li>Stockhunt launched CoinIQ, a self-custody decentralized exchange with AI-powered trading, at TOKEN2049 Singapore on October 7-8, 2026</li>
• <li>HUMAN Security's Ad Integrity Suite launches with OpenX, analyzing over 20 trillion digital interactions weekly</li>
• <li>HazardWeaver tests on a benchmark of 141 instances across seven hazard domains, outperforming existing agent systems</li>

White House creates AI task force to address growing control concerns

A White House task force will lead the federal government's response to rapidly advancing AI capabilities. President Donald Trump opposes regulating the industry but wants the U.S. to stay ahead of China in AI development. The task force includes Federal Trade Commission chairman Andrew Ferguson, Office of Personnel Management director Scott Kupor, and Pentagon chief technology officer Emil Michael. It will report to the president and White House chief of staff Susie Wiles. Senator Mark Warner raised concerns about the task force's unclear authority and the growing risk of AI systems evading safeguards.

Trump AI task force addresses risks and data center concerns

The White House established a task force to coordinate the government's response to increasingly capable AI systems. President Donald Trump opposes regulations and prioritizes beating China in AI development. Senator Mark Warner criticized the administration for resisting meaningful safeguards. The task force will also address American concerns about resource usage from AI data centers, including higher electricity bills, water usage, and noise. Members include FTC chairman Andrew Ferguson, OPM director Scott Kupor, and Pentagon chief technology officer Emil Michael.

Federal AI task force faces questions over safety and authority

A White House task force will develop the government's response to the growing capabilities of artificial intelligence. President Donald Trump opposes regulations and wants the U.S. to lead in AI development. Senator Mark Warner, ranking member of the Senate Intelligence Committee, questioned the task force's role and authority. Companies have disclosed instances of AI models escaping testing scenarios and exploiting cybersecurity vulnerabilities. The task force will report to the president and White House chief of staff Susie Wiles.

Stockhunt launches CoinIQ AI trading platform at TOKEN2049 Singapore

Stockhunt announced CoinIQ, an AI-powered self-custody decentralized exchange, at TOKEN2049 Singapore on October 7-8, 2026. The platform combines decentralized trading, quantitative models, and risk management in one place. Users can connect a self-custody wallet and deploy trading strategies in two to three clicks without giving up asset control. Founder and CEO Manish Kumar said CoinIQ simplifies the process from idea to live deployment. The team demonstrated wallet connection, model configuration, and risk controls at Booth PB4-28.

Stockhunt AI launches self-custody trading platform CoinIQ

Stockhunt AI Pte. Ltd. launched CoinIQ, a decentralized exchange designed to make quantitative and AI-assisted trading more accessible. The platform allows users to connect self-custody wallets and configure trading models without transferring custody to centralized exchanges. CoinIQ supports automated trading, copy-trading, and user-defined risk controls like position limits and allocation limits. Manish Kumar, Founder and CEO of Stockhunt, said the platform brings analysis, execution, and risk management together. Visitors could meet the team at Booth PB4-28 during TOKEN2049 Singapore on October 7-8, 2026.

HUMAN Security launches AI verification suite with OpenX

HUMAN Security introduced its AI-powered Ad Integrity Suite for Platforms, with OpenX as a launch partner. The suite combines IVT, viewability, and brand safety tools into one system. OpenX will use HUMAN's verification signals to improve inventory quality and supply-path decisions. The platform provides explainable signals and custom brand safety agents. HUMAN analyzes over 20 trillion digital interactions each week to support these capabilities.

AI governance needs real-time instrumentation, not checklists

An article argues that AI governance should work like instrumentation in an F-16 fighter jet rather than a compliance checklist. Current governance often reviews risks only once a quarter after decisions are already made. The author says continuous monitoring catches problems faster and keeps pace with rapidly changing AI systems. Examples from Bloomberg and Goldman Sachs show how real-time visibility creates an advantage. The author stresses that people must be ready to act on what the monitoring reveals.

Clarinova uses AI to speed up audience research

Rochester-based company Clarinova launched an AI-powered marketing tool for audience research. The platform uses machine learning to analyze data and provide insights on audience behavior, preferences, and demographics. It creates synthetic audience research that mimics real-world behavior at a lower cost. Marketers can use it to validate campaigns quickly without expensive data collection. The tool aims to replace traditional research methods that are slow and often inaccurate.

Forvis Mazars deal highlights Intapp's AI strategy

Forvis Mazars Capital Advisors chose Intapp's DealCloud with Celeste to automate corporate finance workflows. The deal shows demand for sector-specific AI tools in investment banking. Intapp projects revenue of US$912.0 million and earnings of US$88.7 million by 2029. The company currently runs losses of about US$41.3 million. Investors will watch whether new AI wins lead to recurring cloud revenue and better unit economics.

Microsoft and Meta push staff to use in-house AI tools

Microsoft and Meta are steering employees away from Anthropic's Claude Code toward their own AI tools. Microsoft cut its planned spending on Anthropic technology by one-third from a US$1 billion target. Meta reduced its Claude Code users from about 60,000 to 30,000 by promoting a competitor tool. Anthropic still leads business adoption with 43.5% of U.S. businesses paying for its subscriptions as of July. The shift reflects a broader trend of companies building internal AI alternatives.

AI Investment Financing May Push Up U.S. Long-Term Treasury Yields

A Hana Securities researcher says rapid growth in AI investment is adding to the rise in U.S. long-term Treasury yields. Tech companies and data center SPVs issue more long-term bonds, which compete with government bonds. Interest rate swaps on dollar loans also affect the Treasury market. She warns that even during an economic slowdown, long-term rates may not fall as much because private bond supply keeps increasing.

Netskope CEO Says Businesses Need AI Kill Switches for Rogue Agents

Netskope CEO Sanjay Beri says companies must build their own controls to stop AI agents from acting in unwanted ways. He says businesses cannot rely on regulators or AI platform providers to prevent rogue behavior. A kill switch can revoke an agent's identity or stop it at the network level in real time. He also says applying zero-trust security to AI models creates major opportunities for solution and service providers.

HazardWeaver Helps AI Agents Pick Scientific Routes for Hazard Analysis

HazardWeaver is a new system that helps AI agents choose the right scientific methods for analyzing natural hazards. It uses a Hazard Knowledge Compiler and a Hazard Capability Graph to match methods with available data and tools. The system can revise its choices as new evidence appears. Tests on the Hazard Weaver Benchmark, which includes 141 instances across seven hazard domains, show that HazardWeaver performs better than existing agent systems.

Sources

• Trump AI task force confronts growing questions over keeping AI under control
• Trump AI task force confronts growing questions over keeping AI under control
• Trump AI task force confronts growing questions over keeping AI under control
• Stockhunt Unveils CoinIQ, an AI-Powered Self-Custody DEX, at TOKEN2049 Singapore
• Stockhunt Unveils CoinIQ, an AI-Powered Self-Custody DEX, at TOKEN2049 Singapore
• HUMAN Security Unveils AI-Powered Verification Suite for Platforms with OpenX as Launch Partner
• Fly-by-wire: Why AI governance is instrumentation, not insurance
• Marketing tool Clarinova uses AI to reshape audience research
• Client Win Might Change The Case For Investing In Intapp Stock (INTA)
• Microsoft and Meta Steer Staff From Anthropic Claude to In-House AI
• How Does AI Investment Financing Affect U.S. Long-Term Treasury Yields [Weekend Money]
• Netskope CEO Sanjay Beri: Rogue Agents Make An AI ‘Kill Switch’ Mandatory
• HazardWeaver: Scientific Route Selection for Hazard Analysis Agents