Agents Get Skills. Infrastructure Shifts. Robots Move In.

What happened
Salesforce added 30 new AI features to Slack, expanding the Slackbot agent to handle tasks like drafting emails, scheduling meetings and sorting through inboxes. The update introduces “reusable AI skills” that let users define custom workflows—like building budgets—once and apply them across multiple channels or conversations.

Why it matters
Slack’s shift from simple chat assistant to a flexible work agent shows agentic AI moving into mainstream productivity tools. By letting users create their own skills, Slack bridges bespoke automation and enterprise governance.

What’s next
Expect a marketplace of Slack skills and deeper integrations with Salesforce’s broader platform. As companies adopt these agents, identity and access controls will need to evolve to track who (or what) is acting on behalf of users.

What happened
Oracle announced major layoffs—reportedly around 30,000 employees—while simultaneously ramping up investment in next-generation AI infrastructure.

Why it matters
This signals a dramatic shift in enterprise priorities from headcount to high-performance compute, as companies race to stay competitive in the AI arms race.

What’s next
Expect other tech giants to follow suit, reallocating resources to scale their AI capabilities and infrastructure as costs and stakes rise.

What happened
OpenAI said it was “saying goodbye to the Sora app,” its text‑to‑video generator, acknowledging that the decision was motivated by cost cuts and redirecting resources. Sora cost about $15 million per day to run but generated only $2.1 million in revenue, and downloads plunged after an initial spike.

Why it matters
The demise of Sora underscores the difficulty of finding a business model for flashy generative AI demos. It gives Alphabet’s Gemini and Anthropic’s Claude more breathing room as investors question OpenAI’s heavy spending.

What’s next
Alphabet and Anthropic may capture enterprise customers looking for stable products, while OpenAI faces pressure to turn its research into sustainable revenue. Regulators will watch how AI platforms handle copyright and deepfake concerns without Sora as a foil.

What happened
Governor Gavin Newsom signed an executive order requiring AI companies seeking state contracts to certify within 120 days how they will prevent their technology from being misused. The order directs agencies to develop best‑practice certifications and watermarking guidelines, and mandates that models avoid harmful bias and civil‑rights violations. It comes amid federal scrutiny of Anthropic, which was briefly designated a supply‑chain risk for refusing to tighten guardrails on its AI.

Why it matters
California is using its purchasing power to set de facto standards for AI safety and transparency. By tying contracts to safeguards, the state could influence practices across the industry.

What’s next
Other states and federal agencies may adopt similar procurement rules, while companies scramble to implement watermarking and bias‑mitigation processes. The tug‑of‑war between state orders and a pending national AI framework will intensify.

What happened
Microsoft unveiled plans to invest over $1 billion in cloud and AI infrastructure in Thailand between 2026 and 2028.

Why it matters
This expansion aims to meet surging enterprise demand in Southeast Asia, signaling a global push to localize and scale AI capabilities.

What’s next
Expect increased regional competition and new opportunities for enterprise AI adoption as global players race to establish local AI strongholds.

What happended
UniX AI launched the Panther-series wheeled humanoid robot, now being tested in real homes, capable of cooking, cleaning, and organizing daily routines.

Why it matters
This leap from single-task robots to true household automation brings the vision of a general-purpose home robot closer to reality.

What’s next
Real-world deployments will test the robot’s ability to handle unpredictable home environments, with battery life, cost, and reliability as key hurdles for mainstream adoption.

What happened
IntBot’s humanoid robot “Jose” began a four-month pilot at San Jose International Airport, greeting travelers and answering questions in over 50 languages.

Why it matters
This is a high-profile test of embodied AI’s ability to interact with diverse populations and operate in dynamic, unpredictable public spaces.

What’s next
Success could accelerate adoption of humanoid robots in airports and other public venues, expanding the footprint of AI-powered physical systems in daily life.

AI is shifting from tools you use to systems that act—inside your chat apps, your infrastructure, and even your home—forcing a rethink of control, cost, and accountability across the stack.

Turn your chat app into a real agent:

Open Slack (or any AI workspace you use)
Ask: “Create a reusable workflow to summarize my emails and draft replies”
Then reuse it in a different channel or task

You’ll feel the shift instantly—this isn’t prompting anymore, it’s delegation.