Back to News
Binance now lets AI agents trade, but keeping them in check is largely up to users
AppsThursday, August 20, 2026 at 10:00 AM

Binance now lets AI agents trade, but keeping them in check is largely up to users

Source: TechCrunch AI
Curated Summary

Binance has launched Agent OS, a platform enabling AI agents to autonomously analyze markets and execute trades on behalf of users.

This move integrates major AI tools like OpenAI's ChatGPT, Anthropic's Claude Code, and Cursor with Binance's financial infrastructure, including its APIs and Model Context Protocol support.

The platform utilizes dedicated sub-accounts to sandbox agent activity, blocking withdrawals by default to protect user funds.

While users retain granular control over permissions and can require approval for specific orders, Binance does not impose separate trading caps, relying on the sub-account balance as the limit.

A key security consideration is that Binance cannot see the reasoning behind an agent's decisions, making prompt-injection attacks a potential risk that users must manage.

This launch marks a significant shift in the crypto industry, moving from passive chatbots to active agents managing real money with decentralized responsibility.

Want the full story?

Read Original Article

via TechCrunch AI

#binance#crypto#agentic-ai#trading#mcp

Related Articles

ChatGPT can now send texts for you with new Apple Messages plugin
TechCrunch AI
AppsAug 20

ChatGPT can now send texts for you with new Apple Messages plugin

OpenAI has launched an Apple Messages plugin for ChatGPT, allowing users to connect their Messages inbox to the chatbot. The plugin enables sorting, analyzing, editing, deleting, and drafting/sending messages, and works with Codex and ChatGPT Work for professional use. OpenAI claims the plugin runs locally and doesn't index all messages, but privacy specifics remain unclear. TechCrunch has reached out for clarification. OpenAI also warns users to keep approval settings on to review messages before sending, discouraging persistent approval. This raises privacy questions about AI accessing personal communications.

1 min readRead
Ok, can we actually cool data centers with our pee?
TechCrunch AI
AppsAug 20

Ok, can we actually cool data centers with our pee?

Liquid Death launched a cheeky marketing campaign with former NFL star Jason Kelce suggesting people pee on data centers to cool them, highlighting the massive water consumption of AI infrastructure. While the ad is humorous, experts note that using treated wastewater—which includes urine—to cool data centers is a real and viable approach. Michael Obradovitch of Ecolab confirms alternative water sources are already used to offset potable water demand, and Bruno Pigott of the WateReuse Association explains that wastewater can be cleaned and reused for cooling. However, directly using raw urine would cause mineral deposits and fouling in cooling towers, requiring constant cleaning. The article clarifies that large-scale urine collection isn't practical, but wastewater treatment with membrane bioreactors and reverse osmosis can produce recycled water for data center cooling. This piece ties a viral marketing stunt to a serious sustainability topic relevant to AI's environmental footprint.

1 min readRead
Google gives publishers a new way to fight AI-driven traffic losses
TechCrunch AI
AppsAug 20

Google gives publishers a new way to fight AI-driven traffic losses

Google is rolling out a new 'Preferred Sources' button for publishers to embed on their websites, letting readers mark favorite sources to be highlighted more often in Google Search, Discover, and Google News. The feature expands on an earlier May launch of Preferred Sources in AI Mode and AI Overviews, and Google reports over 345,000 unique sources already selected. Publishers can benefit as users are twice as likely to click through preferred sources, helping offset traffic losses caused by AI-powered search features. Google is also introducing natural-language customization for Discover feeds and audio daily briefings in the Google News app for Android users. This move reflects a broader industry trend of using AI to give users more control over content recommendations.

1 min readRead