📍California Keeps Humans Accountable
Plus: Managers Let AI Write Reviews, Ad Agents Move Into Streaming TV
Good Morning, Agent Owners!
“Human oversight” is starting to sound less like control and more like someone legally required to stand near the robot.
California Keeps Humans Accountable
TL;DR: California has passed three laws limiting AI authority in work, healthcare and education. From July 2027, employers cannot fire or discipline workers based solely on AI, clinicians must retain final judgment over AI recommendations, and EdTech firms cannot train models on student data. The measures reflect California’s broader push to keep humans accountable where automated decisions carry high stakes. Read more →
Managers Let AI Write Reviews
TL;DR: ZDNET reports a Highwire survey of 1,034 corporate employees found 78% of managers used AI to draft, edit or inform performance feedback, yet only 16% of workers were told. While 54% said feedback became more specific and actionable, 34% found it more generic and 32% less useful. Nearly one in four employees also rehearse difficult workplace conversations with AI. Read more →
Haiku 5.5 Has a Catch
TL;DR: Anthropic launched Claude Haiku 5.5 as its fastest small model, with 10¢ input and 50¢ output pricing per million tokens—but only for prompts up to 100,000 tokens; longer prompts get a smaller 50% discount. Anthropic says it beats GPT-6 Luna across six shared benchmarks, while Sonnet 5.5 cache reads were halved. You can now use Haiku 5.5 on MyClaw. Read more →
Amazon Cuts Jobs, Boosts AI
TL;DR: Business Insider reports Amazon cut fewer than 1,000 roles across customer service, marketplace support, retail engineering and overseas teams, following roughly 30,000 job eliminations since late 2025. The reductions come as Amazon prepares a record $220 billion infrastructure push centered on AI, while also recruiting some former employees into AWS and AI roles amid continued workforce restructuring. Read more →
Meta Pushes Agent Commerce Standard
TL;DR: Meta, Walmart, Stripe and Sierra Technologies are backing the Personal Agent Protocol, an open standard meant to authenticate AI agents, protect consumers and give businesses visibility into agent activity. The effort responds to bank concerns over fraud, privacy and biased recommendations, while Meta hopes trusted standards will accelerate adoption of commerce agents like Muse. Read more →
Ad Agents Move Into Streaming TV
TL;DR: AdExchanger reports NBCUniversal, FreeWheel, Carat and Newton are building AI buying agents for streaming TV, combining identity, audience and performance data to recommend inventory against campaign goals. An unnamed luxury retailer is first to test the system. The promise is more outcomes-based CTV buying, though the article warns optimizing every campaign for immediate results could undermine long-term brand building. Read more →
Singapore Sets AI Rules for Finance
TL;DR: Singapore’s Monetary Authority has issued AI risk-management guidelines for all financial institutions, covering oversight, lifecycle risks and third-party systems as AI becomes more autonomous. Firms can tailor controls to their risk profile, but must assign clear accountability and manage AI proportionately. The rules phase in from October 2027, with full implementation required by October 2028. 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👇
