Coinbase Opens Agent API for Automated Trading and Research Payments
On Thursday, June 11, 2026, the AI landscape saw a mix of developments, from new product releases to controversy and policy reversals. Coinbase has launched an agent API for automated trading and research payments, allowing users to delegate crypto trading and pay for premium research data without manual intervention.
In funding news, Anthropic closed a significant deal, though details are scarce. Meanwhile, Anthropic CEO Dario Amodei is advocating for mandatory pre-deployment audits for frontier AI models, proposing a regulatory framework that would grant governments the authority to block or reverse the release of models that fail independent safety audits.
On the infrastructure side, a study using NASA satellite thermal imagery found that AI data centers are contributing to localized warming, with temperatures rising up to 9°C in areas surrounding these facilities. This environmental impact is a concern as the hyperscale buildout continues to expand.
In a post, Simon Willison argued that Anthropic's decision to reverse its stealth throttling policy on Claude Fable 5 was a necessary step after the AI research community discovered hidden guardrails that silently degraded responses from users engaged in frontier AI research. Willison noted that the move was a response to backlash from researchers who felt the policy could have sabotaged their work.
Anthropic is also partnering with Tata Consultancy Services to distribute its AI models to enterprises, a move that could accelerate AI adoption in the business sector. The partnership will involve TCS spinning up a dedicated Anthropic practice to help companies integrate AI into their operations.
OpenAI's threat intelligence team has published a report detailing an influence campaign that used ChatGPT to generate synthetic content aimed at shaping U.S. public opinion on AI and data centers. The campaign highlights the potential for AI to be used in disinformation efforts.
Simon Willison released datasette 1.0a33, a significant step towards a stable 1.0 release, extending the ?_extra= pattern to cover queries and rows in addition to tables. He also released asyncinject 0.7, a utility library that supports an asyncio dependency injection pattern, which was used with Datasette and helped identify bugs in the dependency.
Willison also announced the release of datasette-agent 0.2a0, which includes new features such as the ability for tools to ask users questions mid-execution. This release enables more interactive and dynamic conversations with the agent.
In a statement, OpenAI expressed support for the EU Code of Practice on AI content transparency, aiming to advance provenance standards and tools to help people understand AI-generated content.
That's the Daily AI Brief for Thursday, June 11, 2026. More tomorrow.
Sources covered
12 stories · 12 sources (7 news, 1 company, 4 voices)
- Coinbase Opens Agent API for Automated Trading and Research Payments📰TechCrunch↗
- Anthropic CEO Pushes Mandatory Pre-Deployment Audits for Frontier AI Models📰The Hill↗
- Anthropic Reverses Stealth Throttling on Claude Fable After Researcher Backlash📰The Verge↗
- Anthropic Bets on TCS as Enterprise Distribution Pipeline📰TechCrunch↗
- AI Data Centers Are Running Hot: Thermal Mapping Shows Localized Warming of Up to 9°C📰Al Jazeera↗
- How a Chinese Influence Campaign Used ChatGPT to Stoke Data Center Backlash📰Engadget↗
- Anthropic Reverses Stealth Degradation Policy for Claude Fable 5 After Community Backlash📰Wired↗
- datasette 1.0a33
- asyncinject 0.7
- Anthropic Walks Back Policy That Could Have ‘Sabotaged’ AI Researchers Using Claude
- datasette-agent 0.2a0
- Supporting Europe’s work in ensuring a trustworthy AI ecosystem