TokenPost topic
AI Agents
511 TokenPost stories are tagged AI Agents. The newest appear first.
Latest AI Agents stories
- Global AI Agent Market Seen Reaching $42B-$53B by 2030
Five research institutions project the market will reach $42 billion to $53 billion, while current use remains focused on market queries and monitoring.
· Technology
- Celo Core Co. Launches buy Marketplace for AI Agent Payments
The marketplace lists more than 700 computing, data and API services that agents can pay for with USD Coin, USD₮ and USA₮.
· Technology
- Nvidia Brings Local AI Agent Workshop to Paris on Oct. 13
The sold-out Build-A-Claw session features OpenClaw, Hermes and physical-AI demonstrations on Nvidia-accelerated hardware.
· Technology
- American Express Unveils Framework for Authenticated AI-Agent Purchases
Intent Intelligence records customer-approved limits and creates scoped payment credentials without exposing card numbers to AI agents.
· Technology
- Meta and OpenAI Target Different AI-Agent Customers as Markets Overlap
Meta’s Muse focuses on consumer tasks, while OpenAI’s Dots serve professionals and teams. Both companies are expanding beyond their initial customer focus.
· Technology
- Consumer Internet Sector Median Falls 11% as AI Risks Weigh
The sector’s median valuation decline contrasts with a 14% gain for the S&P 500 this year.
· Investing
- On-Chain Security Incidents Set Monthly Records as AI Agent Risks Rise
Blockaid co-founder Ido Ben-Natan said attacks on blockchain infrastructure are increasing as AI agents gain access to more funds.
· Technology
- Coinbase and Nansen Expand AI Tools for Crypto Trading Access
Coinbase enables agent-driven trading within user-defined limits, while Nansen requires final confirmation before transactions execute.
· Technology
- Nansen CEO Predicts AI Agents Will Handle Crypto Trades by 2028
Robinhood has opened more than 150,000 agentic accounts, while failed automated trade attempts highlight growing security risks.
· Technology
- Enterprise AI Governance Moves From Policy To Runtime Controls
Companies are adding controls to monitor agent actions, data access and tool use as autonomous systems enter cloud, endpoint and business workflows.
· Technology
- Meta Launches Muse Wonder Models for Coding and Agentic Work
Meta says Muse Wonder and Muse Wonder Full deliver frontier-level performance, with Wonder available through the company’s Model API.
· Technology
- t54 Documents x402 Payment Tools for AI Agents on XRP Ledger
The implementation supports XRP and Ripple USD (RLUSD) for machine-to-machine payments tied to APIs, data retrieval and AI inference.
· Technology
- Meta Expands Muse AI Agent Alongside $130B-$145B 2026 Spending Forecast
The personal AI agent is moving into shopping, enterprise software, developer tools and AI glasses, while Meta has not disclosed Muse-specific revenue or profitability.
· Technology
- Mastercard Expands AI Agent Controls for Automated Payments
The company is testing a transaction score in the United States while liability for mistaken purchases within approved limits remains unresolved.
· Technology
- OpenAI, Anthropic Disclose AI Agents Bypassing Safeguards
Claude submitted a fabricated homicide tip to a Philadelphia police website, while OpenAI models accessed external systems during internal evaluations.
· Technology
- Andrej Karpathy Estimates 75% Barely Contact Large Language Models
His rough four-level estimate ranges from 1 billion to 2 billion casual users to about 5,000 people with internal access to frontier-grade systems.
· Technology
- UC San Diego Researchers Train AI Agents to Model Partners’ Beliefs
The framework improved results across language and multimodal benchmarks, including an 11.5% relative gain on Sotopia and a 26.7-point increase on BigToM.
· Technology
- Sui and Teranode Show Two Paths to High-Throughput Settlement
A Sui test reached 40,614,180 TPS inside off-chain tunnels, while Teranode sustained 1 million TPS across six cloud regions in a separate benchmark.
· Technology
- Meta’s Muse Grew From OpenClaw Experiments Into a Personal AI Agent
The agent launched in the United States on Sept. 8 after about seven months of reliability testing and product development focused on everyday digital tasks.
· Technology
- Grayscale Research Chief Says Cheaper AI May Increase Compute Demand
Zach Pandl said lower token costs and more autonomous software could expand AI usage, increasing demand for GPUs, data centers and power.
· Technology
- AI Agents Face Security Risks That Can Outweigh Their Benefits
Testing found successful attacks against all 13 evaluated frontier models, while broader access to data and tools increases the consequences of mistakes.
· Technology
- GitHub Rebuilds Storage Layer as September Commits Top 7.38 Billion
The redesign will move authoritative repository data to Azure Blob Storage and separate durable storage from request-handling compute.
· Technology
- Grok Bot Uses Persistent Cloud Computers to Keep AI Agents Working
The beta system gives each user one persistent Firecracker microVM in the United States, allowing Bots to operate across tools after a laptop is closed.
· Technology
- Trump AI Policy Leaves Liability Questions as Congress Targets Agent Harm
The administration favors federal uniformity and voluntary safeguards, while lawmakers propose civil and criminal liability for companies behind harmful AI agents.
· Regulation
- Stripe Co-Founder Says AI Agents Could Make Product Quality Matter More
Patrick Collison said personal AI agents could compare prices, discounts, reviews and after-sales service, reducing the advantage of advertising and distribution.
· Technology
- Grok Bot Adds Dedicated Email Address for Service Sign-Ups and Scheduling
The SpaceXAI agent can use the address to register for services, contact merchants and arrange schedules for users.
· Technology
- Citrini Research Shifts Crypto Thesis Toward Blockchain Infrastructure
The Oct. 8 analysis focuses on tokenized assets, programmable finance and artificial-intelligence agents operating across on-chain applications.
· Technology
- Enterprise Willingness to Let AI Agents Make Production Changes Falls to 56%
The August survey found 42% of organizations deploying autonomous agents expected to retain human review, while evaluation-platform adoption continued to rise.
· Technology
- HSBC And Ant Digital Test AI-Agent Payments With Tokenized Deposits
The test used HSBC’s Tokenised Deposit Service, Ant Digital Technologies’ Anvita Flow network and Jovay Testnet for real-time settlement and risk checks.
· Technology
- Microsoft Makes Windows 11 a Hub for Local and Cloud AI Agents
Execution Containers now control what agents can access, while Microsoft prepares local Copilot features and larger on-device models.
· Technology
- Aave MCP and MetaMask Agent Wallet Serve Separate AI DeFi Roles
Aave’s service provides protocol data and unsigned transactions, while MetaMask adds custody controls, spending limits and transaction-security features for AI agents.
· Technology
- Instinct Faces New Test as Meta and OpenAI Expand AI-Agent Access
The invite-only assistant raised $1 billion at a $10 billion valuation as Muse and Dots bring browser-capable agents to broader platforms.
· Technology
- AI Agents Make Company Documentation an Operational Layer
Mintlify documentation sites recorded 257 million agent requests in August 2026, compared with 131 million human page loads, as developers reported persistent concerns about AI accuracy.
· Technology
- AI Agents Can Pay for Shopping as Refund Rules Remain Unclear
Amazon’s AgentCore Payments gives AI agents wallet access, spending limits and transaction monitoring, but dispute responsibility varies by payment method and platform.
· Technology
- AI Agents May Favor Costlier Options Based on Wealth Signals, Study Finds
A preprint covering 325,000 experiments found that eight of 13 models often selected pricier fixed-price options for wealthier synthetic users. The research did not test crypto wallets or agents.
· Technology
- Zardot CTO Says More Than 80% of Software Workers May Face Career Crisis
Obie Fernandez said AI tools can handle requirements, coding, testing, documentation and Jira tasks, increasing the importance of judgment and problem selection.
· People
- Coinbase and Robinhood Build Infrastructure for Agentic Finance
Autonomous software could manage funds and transactions while tokenized assets trade around the clock, directing attention toward platforms that capture fees.
· Technology
- Elastic Launches AlertZero AI Agents for Security Alert Triage
The technical preview divides security work among four specialized Watches and lets customers set approval requirements for different tasks.
· Technology
- AWS Adds Infrastructure and Governance Tools for AI Agents
The releases span runtime performance, managed agents, architecture reviews and sandbox controls, with features at different stages from general availability to developer preview.
· Technology
- OpenAI Board Member Urges Layered Safeguards as AI Agents Gain Access
Zico Kolter said scaling alone will not protect AI systems from manipulation as agents gain access to data, software tools and sensitive systems.
· Technology