📍Uber Gives 30,000 Staff AI Help
Plus: Agents Take the Night Shift, Nestlé Speeds Up Retailer Replies
Good Morning, Agent Owners!
AI now works at Uber, handles legal cases, and shops for groceries. Congratulations, humanity has been promoted to the approval button.
Uber Gives 30,000 Staff AI Help
TL;DR: Uber is expanding an AI agent rollout from an initial 1,000-person pilot to availability for more than 30,000 employees worldwide. Through approved connections to workplace tools, the agent can research information, analyze data, and prepare finished work across multiple steps. Internal training and a skills library support the expansion; the announcement gives no active-user count or measured productivity gains. Read more →
Agents Take the Night Shift
TL;DR: At Rescale, an AI agent investigates nightly test failures by reading logs, checking historical incidents, and examining code before recommending fixes for engineers to review. The company reports a 20% reduction in deployment delays while failures are analyzed. Its classifications agree with human triage 68% of the time, making this an automated first pass that engineers continue to validate. Read more →
Five Days of Pricing, Two Hours
TL;DR: Restaurant owner Ted Taylor uses an AI agent to compare ingredient costs, competitors’ prices, and sales patterns before producing proposed menu prices for his three locations. He says pricing analysis that once took five days now takes one to two hours. Taylor reviews and adjusts the recommendations himself; the figures describe one operator’s experience, detailed in a new customer case study. Read more →
Legal Case Checks Go Parallel
TL;DR: Broughton Partners has deployed 18 AI agents to check identity, fraud indicators, contracts, and medical evidence before cases reach law firms. Six agents can assess medical records in parallel, recording evidence and reasoning for reviewers. Its five-person team handles ambiguous findings and retains final decisions. The agents never reject claimants, keeping human judgment inside an increasingly automated audit workflow. Read more →
Nestlé Speeds Up Retailer Replies
TL;DR: Nestlé is piloting an agent that reads retailers’ emails and attachments, finds the relevant account and order information, and drafts replies for staff approval. Order-information searches fell from several minutes to under one minute in pilot scenarios. Early tests suggest more than 55% automation potential in selected service interactions, with employees still reviewing responses and handling requests requiring additional judgment. Read more →
Walmart Welcomes Automated Shoppers
TL;DR: Walmart has opened a beta gateway for external AI agents to shop on customers’ behalf. Agent profiles planned for next month will let shoppers separate agent purchases, require approval, set spending limits, and revoke access. Purchases require customer approval by default, and agents cannot change payment methods. The retailer is rolling out these controls in stages as automated shopping grows. Read more →
The Checkout Page Still Matters
TL;DR: AI-agent orders on Forter’s merchant network have grown to roughly five times their initial level, according to data spanning July 23 to October 20. Agents operating merchants’ checkout pages accounted for 70% of agent orders in the week of October 12. The findings describe Forter’s network, whose absolute sample size is undisclosed, rather than the wider e-commerce market. Read more →
Finding Bugs Is Just the Start
TL;DR: AI-assisted vulnerability discovery has triggered a coordinated response to critical PHP flaws involving maintainers, researchers, infrastructure providers, and government agencies. OpenAI says it spent the past two months helping organize that response, while PHP maintainers led fixes and releases. The episode shows how finding a serious bug creates a second workload: validating patches, preparing protections, and getting updates to affected organizations. Read more →
The Surprise Is on the Bill
TL;DR: Unexpected cloud costs were the most common negative consequence of ungoverned agents in a Redwood-commissioned survey of 300 technology, risk, and AI leaders and practitioners. Seventy-seven percent of respondents reported some negative consequence, with operational disruption and compliance issues also cited. The vendor-funded findings highlight deployment costs and ownership gaps, while describing this respondent sample rather than enterprises as a whole. 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👇
