Industry

AI News Flash: A study called ‘Emergence World’ lets 10 AI agents live autonomously in a virtual city for 15 days to verify if short-term testing can assess AI’s long-term behavior risks. Researchers point out that the industry currently uses an ’exam mode’ for testing AI agents—giving a single task in a clean environment and drawing conclusions within minutes. But real-world autonomous systems often need to run for weeks or even months, interacting with other AIs whose behavior isn’t controlled by a single operator…

AI News Flash: AI agents (Agentic AI) are breaking traditional boundaries between artificial intelligence and financial transactions. Previously, AI systems could only answer questions, organize files, or assist with programming - critical operations like account access and transfer confirmations were always performed by humans. Now, new autonomous agent architectures allow AI to set goals, call external tools and execute tasks on its own. Developers are actively exploring connecting these agents to crypto wallets. Once integrated with on-chain infrastructure, application areas will far exceed simply querying token prices. In theory, agents can monitor DeFi portfolios 24/7, automatically organize and rebalance required transactions, pay for digital services, search for yield opportunities, and even execute predefined financial commands when users are offline. The blockchain’s global openness, continuous operation, and programmable nature make it an ideal track free from business hours and geographic limitations. Notably, the first widespread crypto application of AI may not be trading but forming a new economic model where software purchases services from software - paying for APIs, cloud computing, and data set licenses directly.

Grayscale researcher Pandl argues Bittensor represents a decentralized AI alternative akin to Bitcoin for open global AI access. As AI capabilities grow, governments and labs increasingly control who can access these powerful tools.

AI news flash: The US government issued an emergency directive requiring Anthropic to remove its new Claude models (Fable 5, Mythos 5) from public access by late Friday. Anthropic stated in their official blog that the directive stemmed from a government misunderstanding of the threat level posed by a ’non-universal jailbreak’ vulnerability, which originated from an anonymous report. Claude currently has approximately 18,900 monthly active users, and Anthropic says it’s actively fixing security issues to restore user access as quickly as possible.

AI News Flash: Anthropic officially released the first public version of the Mythos model series, Claude Fable 5. The company revealed last month that over 10,000 high-severity or critical vulnerabilities targeting systemically important software were discovered during Mythos testing, sparking widespread debate over whether the model should be publicly deployed. Anthropic decided to release it…

AI News Flash: Coinbase launches in-app AI agent ‘Coinbase Advisor’ registered as a licensed financial advisor with US SEC and CFTC, capable of automatically executing recurring purchases and fund allocation tasks without continuous human monitoring. Using ETH dollar-cost averaging as example, the agent automatically fetches 30 days of hourly price data, identifies the lowest historical price window, and sets up $20 daily market purchases for two consecutive weeks…

AI News Flash: Immunefi CEO Amador warns that the next three to four years will be a make-or-break period for the crypto industry. He says the whole industry will keep facing AI-accelerated infiltration until security teams can put AI’s defensive capabilities to work and build codebases attackers simply can’t crack. If the industry adopts crowdsourced security more broadly, letting more researchers turn AI into a defensive tool, this danger window could shrink.

AI News Flash: BEAT, the AI token from Audiera, has surged over 1,500% in the past month, massively outperforming Bitcoin and Ethereum. Three main forces are driving this rally. First is the platform’s revenue-burn mechanism: between June 1-8, Audiera reported weekly revenue of 772,045 BEAT tokens (about $2.87M at $3.712), while burning 770,545 tokens over the same period.

AI News Flash: Anthropic’s newly released Fable 5 model has drawn heavy criticism since launch, with the core controversy centering on its overly strict safety guardrail design. When users ask about sensitive topics like bioweapons or cybersecurity, the model doesn’t just return a warning — it automatically downgrades and routes the conversation to an older, less capable model, marking the first time the AI industry has used a ‘downgrade routing’ mechanism to handle sensitive queries. Princeton AI researcher Sayash Kapoor says this is a rare case of a guardrail launch getting unanimous backlash across the industry.

AI News Brief: Tether has been ramping up its AI push over the past few months, built around its in-house QVAC platform. Back in March, Tether launched a training framework that lets AI models train and run inference directly on consumer hardware, including smartphones and non-Nvidia chips, breaking away from the old reliance on high-end GPU servers. Two months later, its QVAC MedPsy medical AI model…