Meta is moving forward with its AI agent ambitions despite growing skepticism. The company launched a personal AI agent called Muse, but analysts question whether it will gain traction. Meta's track record on data safety remains a concern for users who need to trust these agents. The company recently agreed to pay $18 billion to settle lawsuits related to harming children, and its rich stock valuation adds pressure to take bold risks on AI strategy.
On a more collaborative note, Meta is joining forces with Walmart and Sierra to develop tech standards for doing business with AI bots. The effort aims to bring order to the chaos companies face as AI agents enter the workplace. The new standards would help organizations distinguish personal agents from real people and improve security measures. OpenAI and Anthropic have not yet signed on but are expected to participate.
Meanwhile, investor Vinod Khosla is backing Wajo, an AI agent developed by ex-DeepMind engineer Poddar. Wajo's Fo agent competes with offerings like Muse and Instinct but distinguishes itself through a focus on trust and data privacy. The agent can call businesses, create virtual credit cards, and hire humans when needed. Wajo has grown 10x since its late September launch and is now used across 107 countries.
In government news, Jay Clayton, Director of National Intelligence, is leading President Trump's Super Intelligence Force, a task force designed to centralize AI policy oversight within the White House. Scott Kupor, co-chair of the task force, defended a flexible approach to AI regulation, emphasizing the need for adaptability in both development and oversight.
TikTok has entered the commerce space with a new AI-powered Shopping Assistant and in-app checkout feature. The tool provides real-time guidance on product details, shipping, sizing, and availability. It was built in partnership with Salesforce, Shopify, Shoplazza, and Stripe. TikTok Shop, which launched in 2023, has already generated $2.08 billion in beauty and personal care sales.
Security concerns around AI agents are mounting. Researchers found hidden prompt injections in about 1% of roughly 200,000 résumés, showing attempts to trick AI hiring screeners. North Korean IT workers have also used AI and face-swapping to conceal identities during remote interviews. Experts recommend repeatable verification to confirm that the person who interviewed is the same person who ends up working at the company.
On the hardware and infrastructure front, Vicor Corporation's stock nearly doubled in September as demand for its AI power technology surged. The company's one-year backlog reached $379.7 million, up 26% from the previous quarter, with four major customers licensing its Vertical Power Delivery technology. However, its first chip fab is nearing capacity, and a second fab in New Hampshire will take about a year to deploy.
Binance also entered the agent space with Binance Intelligence, a suite that includes an AI trading agent called Binance AI Pro. Users type requests in plain language and receive a strategy with a visual flowchart they can adjust before approving any trade. Live trading requires a Premium plan costing 19.99 USDC per month. The agent has been in beta since March 25, and Binance plans to offer a free version to all 300 million registered users.
Key Takeaways
- Meta launched AI agent Muse but faces analyst skepticism over data safety and a $18 billion settlement for child harm lawsuits.
- Meta, Walmart, and Sierra are creating AI bot business standards, with OpenAI and Anthropic expected to join.
- Vinod Khosla backs Wajo AI agent, which has grown 10x since late September and operates in 107 countries.
- Jay Clayton leads Trump's Super Intelligence Force, with Scott Kupor advocating a flexible AI policy approach.
- TikTok launched an AI Shopping Assistant built with Salesforce, Shopify, Shoplazza, and Stripe, as TikTok Shop hit $2.08 billion in sales.
- About 1% of 200,000 résumés contained hidden prompt injections targeting AI hiring screeners, researchers found.
- Vicor's AI power technology backlog reached $379.7 million, up 26% quarter-over-quarter, but fab capacity is constrained.
- Binance introduced Binance AI Pro, a plain-language trading agent requiring a 19.99 USDC monthly Premium plan for live trading.
- 61% of UK workers received no training on AI-driven cyber threats in the past 12 months.
- Chinese humanoid robots demonstrated before South Korea's National Assembly, highlighting a 30 billion yuan annual Chinese robotics investment.
Skepticism grows over Meta's AI agent strategy
Meta launched a personal AI agent called Muse, but some analysts doubt it will succeed. Users need to trust AI agents, yet Meta has a poor record on data safety. The company recently agreed to pay $18 billion to settle lawsuits about harming children. Rich stock valuation also raises concerns about the risky AI strategy.
Meta and Walmart join effort to set AI bot standards
Meta, Walmart, and Sierra are creating tech standards for doing business with AI bots. The goal is to bring order to the confusion companies face when AI agents enter the workplace. The new standard would help companies tell personal agents from real people and improve security. OpenAI and Anthropic are not yet on board but are expected to join.
Vinod Khosla backs Wajo AI agent for trust focus
Investor Vinod Khosla believes Wajo, an AI agent built by ex-DeepMind engineer Poddar, will win the agent market through trust and safety. Wajo's Fo agent works like rivals Muse and Instinct but focuses on hiding user data and building trust from day one. The agent can call businesses, create virtual credit cards, and even hire a human when needed. Wajo has grown 10x since its late September launch and is used in 107 countries.
Trump's AI co-chair backs flexible approach
Scott Kupor, co-chair of President Trump's super intelligence task force, defended a flexible AI strategy. The task force aims to guide national AI policy. Kupor spoke about the need for an adaptable approach to artificial intelligence development and regulation.
Jay Clayton takes charge of Trump's AI task force
Director of National Intelligence Jay Clayton is leading President Trump's Super Intelligence Force. The task force is designed to bring AI policy oversight together in the White House. Clayton is seen as a serious and competent pick but will face challenges in mediating disagreements between Cabinet officials.
AI Interview Bots Face Attacks and Can Be Manipulated
AI tools used in hiring can be both targets and tools of attack. Researchers found hidden prompt injections in about 1% of roughly 200,000 résumés, showing attempts to trick AI screeners. North Korean IT workers have used AI and face-swapping to hide identities during remote interviews. Experts recommend repeatable verification to confirm that the person who interviewed is the same person who works there.
TikTok Adds AI Shopping Assistant and One-Click Checkout
TikTok launched an AI-powered Shopping Assistant and in-app checkout to help users buy products without leaving the app. The assistant offers real-time guidance on details, shipping, sizing, and availability. It was built with partners including Salesforce, Shopify, Shoplazza, and Stripe. TikTok Shop, launched in 2023, has generated $2.08 billion in beauty and personal care sales.
Chinese Humanoid Robots Showcase at South Korea Assembly
Chinese humanoid robots appeared before South Korea's National Assembly committee, greeting lawmakers and performing demonstrations. The robots were remotely controlled and showed some awkward moments during the display. A lawmaker highlighted that China invests about 30 billion yuan yearly in robotics, while South Korea has only 53 hardware companies, over half of which have no external investment. South Korea's Science and ICT Minister promised to make physical AI a key growth focus.
Most UK Workers Untrained on AI Cyber Threats
About 61% of UK workers received no training on AI-driven cyber threats in the past 12 months. Over 22% said their organization has no policies against phishing, deepfake voice scams, or AI-generated fake invoices. Only 25% received specific training on AI threats, while 30% had AI threats covered only in general cybersecurity training. Younger workers and non-management staff showed the lowest confidence in spotting these threats.
HPE and Dell Compete in AI Hardware Market
Dell and HPE reported strong AI-related results but differ in key areas. Dell raised fiscal 2027 revenue guidance to $192 billion and expects $74 billion in AI server revenue, with a $95 billion AI backlog. HPE achieved a 40% gross margin but faces supply limits as networking orders grow 3.5 times faster than revenue. Dell leads on operating margin, revenue growth, and lower debt, while HPE is slightly cheaper against sales.
North Dakota AI and Data Center Committee discusses Harmony township data center proposal
The North Dakota Legislative Council AI and Data Center Committee met at Grand Farm in Wheatland on October 6, 2026. Lawmakers heard from industry experts about how data centers impact the state. A major data center proposal in Harmony township near Casselton and Mapleton has sparked public debate. Residents feel left out of decisions about a large industrial project in their area. Some speakers called for clearer state guidelines on data center development.
Vicor sees strong AI power demand but faces scaling challenges
Vicor Corporation's stock nearly doubled in September as interest in its AI power technology grew. The company's one-year backlog reached $379.7 million, up 26% from the previous quarter. Four major customers have licensed Vicor's Vertical Power Delivery technology. However, the company's first chip fab is nearing capacity, and a second fab in New Hampshire will take about a year to deploy. The stock's forward P/E of 47.17x leaves little room for error as competitors like Monolithic Power Systems and Texas Instruments compete in the same space.
Binance launches AI agent that builds trading strategies from plain language
Binance introduced Binance Intelligence, a suite including an AI trading agent called Binance AI Pro. Users type requests in plain words and receive a strategy with a visual flowchart that they can adjust before approving any trade. Live trading requires a Premium plan costing 19.99 USDC per month, launching in the second half of October. The agent has been in beta since March 25. Co-CEO Richard Teng and VP of Product Jeff Li presented the suite during a livestream, and Binance plans to offer a free version to all 300 million registered users.
Sources
- Why I Don't Think Meta's AI Strategy Will Be Successful
- Meta joins with group of companies to tame ‘chaos’ of doing business with AI bots
- Vinod Khosla believes ex-Deepmind engineer's Wajo will win agent market on trust
- Trump's AI co-chair defends flexible AI strategy
- AI & Tech Brief: Jay Clayton takes control
- AI Interview Bots Are Now Both the Attacker and the Target
- TikTok introduces AI shopping assistant and one-click checkout for in-app purchases
- Chinese Humanoid Robots Appear at National Assembly Audit... "South Korea's Physical AI Investment Falls Woefully Short"
- Spooky! Nearly 2/3s of UK Workers Have No Training on Spotting AI Cyebr Threats
- Betting On AI Hardware: Is HPE Or Dell The Better Way?
- AI and Data Center Committee meets at Grand Farm as discussion continues over Harmony township proposal
- Vicor (VICR) Has the AI Power Opportunity. Can It Actually Scale?
- Binance AI Agent Builds Strategies From Plain Words, Charges 19.99 USDC To Trade
Comments
Please log in to post a comment.