📍New Models Arrive
Plus: Startups Stay Tiny With AI, Musk Says AI Could Master Fields
Good Morning, Agent Owners!
The agent economy has a simple pitch now: fewer employees, cheaper models, more permissions.
OpenAI and Anthropic Cut Model Prices
TL;DR: Anthropic and OpenAI released cheaper models, their first launches since industry leaders called for slowing advanced AI development. OpenAI introduced GPT-6 Sol and Luna at roughly half GPT-5.6’s previous API pricing, while Claude Opus 5.5 costs about 40% less than Opus 5. All three new models are now available on MyClaw. Read more →
Muse Pushes Privacy Boundaries
TL;DR: Meta’s Muse agent is drawing privacy concerns after users reported highly personal suggestions and unexpected access to messages and habits. One tester said Muse referenced texts he never knowingly shared, while Meta said the agent only accesses data with permission. Others found it repeatedly encouraged linking bank accounts, email, IDs and personal documents to expand its long-term memory. Read more →
Startups Stay Tiny With AI
TL;DR: WSJ reports startups are using AI to grow with far fewer employees, automating work once handled by engineers, marketers and operations staff. Butternut AI cut its team from nine to four while serving 10,000 paying customers, while other startups are reaching multimillion-dollar revenue with small teams. Read more →
Kingfisher Turns AI Into Sales
TL;DR: European home-improvement retailer Kingfisher, owner of B&Q and Screwfix, says AI-driven personalization grew 16% and generated about £100 million in sales, supporting its digital business as it raised its full-year profit outlook. The company is also using AI for content, SEO and conversion, while planning natural-language search and agent-enabled shopping as e-commerce reaches 22% of group sales. Read more →
Best Approved Is Not Best Agent
TL;DR: Australia’s Actuaries Institute warns companies may choose AI agents because they pass cybersecurity and procurement reviews easily, not because they perform the job best. That can leave businesses with compliant but underpowered tools, weaker outcomes and more downstream errors at scale. The report says firms should start with business needs, then build governance and quality controls around the actual risks. Read more →
AI Shopping Moves Toward Checkout
TL;DR: Global Payments says consumers now expect AI agents to make 15% of their purchases within five years, up from 9% a year ago. Nearly half of Americans have used or would consider an AI shopping agent, with comfort rising sharply for everyday purchases. But trust remains the constraint, as 50% still worry about payment security and 42% fear wrong purchases. Read more →
Musk Says AI Could Master Fields
TL;DR: Elon Musk says AI could “beat all fields” by the end of 2027 or 2028, accelerating an earlier prediction by researcher François Chollet that AI would become central across disciplines such as physics, chemistry and biology. Musk’s claim extends his increasingly aggressive AI timelines, though the article notes today’s scientific AI tools remain far from independently mastering entire fields. Read more →
Your agent cloud, everywhere — in your chat apps, browser, and tools. Always on.
- Add to Slack, Teams, Telegram or WhatsApp.
- With a clean, redesigned WebUI.
- Connect any of 100+ apps & sources in one click.
- GPT 6 Astra, DeepSeek v4.1 & more frontier AI models, built in at discounted rates.
Run your agent on MyClaw.ai in Seconds👇
