IT a.m. 🌤️ Apr 23, 2026

🔥 Lead Story

Microsoft Copilot Agentic AI Now Live by Default in M365 Apps

What’s Happening

Microsoft has made Copilot’s agentic features generally available in Word, Excel, and PowerPoint, enabled by default for Microsoft 365 Copilot and Microsoft 365 Business Premium subscribers. According to Microsoft, these agents can now autonomously perform multi-step tasks like drafting documents, analyzing data, and building presentations with minimal user prompting.

The Details
  • According to Microsoft’s announcement, agentic capabilities are on by default, meaning IT teams must proactively review governance policies, data access controls, and acceptable use guidelines before agents act on enterprise data.
  • The rollout covers Microsoft 365 Copilot licensed tenants globally. Admins can manage agent availability through the Microsoft 365 admin center and Copilot controls.
  • Security leaders should assess how autonomous AI actions interact with existing DLP, compliance, and sensitivity label policies, as agents can now read, summarize, and generate content across connected workloads without step-by-step user approval.
Lead Story

🛡️ Cybersecurity News

Anthropic is investigating reports of unauthorized access to Claude Mythos, its next-level AI model capable of enabling dangerous cyberattacks. A small group of users in a private forum gained access on the same day Anthropic launched the model.

Researchers flagged compromised npm packages delivering a self-propagating worm that spreads through stolen developer npm tokens.

📈 M&A News

Vast Data raised approximately $1 billion in Series F funding at a $30 billion valuation amid rising demand for AI infrastructure. The round reflects accelerating investment in infrastructure to support artificial intelligence workloads.

🏗️ Infrastructure News

Google introduces new TPUs that are faster and cheaper than previous versions to compete with Nvidia. The company continues supporting Nvidia chips in its cloud infrastructure.

TSMC will hold off on deploying ASML’s most cutting-edge lithography machines for chip production through 2029 to save money. This could be a setback for the Dutch equipment maker.

🧠 AI News

Salesforce expanded its Google Cloud partnership to connect CRM data with productivity tools for AI agents. The integration aims to break down data silos between enterprise systems.

🌩️ Cloud News

Google Cloud unveiled Agentic Data Cloud infrastructure platform to act as a central nerve center for AI agents. The platform aims to enable unrestricted data access for autonomous artificial intelligence agents.

⚙️ DevOps News

GitHub suspended new subscriptions for several Copilot tiers following surge in demand from agentic coding workflows that consume significantly more compute resources than earlier AI assistance models.

🧠 Editor’s Take

Google Bets Its Cloud Future on Becoming the AI Agent Control Plane

What’s Happening

According to SiliconAngle, Google used Cloud Next 2025 to position Gemini not as a standalone model but as the connective tissue for enterprise AI agents – a strategic play to own the emerging agent control plane where orchestration, governance, and multi-agent coordination happen.

The Details
  • According to SiliconAngle, Google launched the Agent Development Kit and Agent2Agent protocol, aiming to set the standard for how AI agents communicate across platforms – a direct challenge to Microsoft and AWS for infrastructure-layer dominance.
  • The report notes Google is embedding Gemini across its full stack – from Workspace to security to data analytics – making the model an integration layer rather than a product, which raises switching cost implications IT leaders should evaluate now.
  • Per the analysis, this mirrors the cloud platform wars of the 2010s: whoever controls agent orchestration and governance will likely control enterprise AI spending for years. IT decision-makers should assess multi-cloud agent strategies before lock-in deepens.
Editor's Take

🔎 Dive Deeper

CIOs must decide whether to adopt AI-native startups or wait for existing vertical SaaS platforms to catch up. AI is transforming enterprise software decision-making for technology leaders.

CIOs face challenges as AI success requires strong data management practices, but funding for data transformation has been limited. With AI now a CEO priority, addressing accumulated data debt is critical.

Sovereign architectures are emerging as the defining framework as enterprises rethink control over their entire technology stack. The shift is driven by regulatory pressure and geopolitical uncertainty reshaping technology procurement.

LinkedIn’s effectiveness for MSP lead generation is questioned as the platform has evolved since 2017. The article examines whether LinkedIn remains a viable marketing channel for MSPs.

The CEO behind Apple’s storied trillion-dollar run bet big on business customers. The key to maintaining that foothold is AI.

B2B buyers now expect vendors, especially AI product providers, to prove solutions deliver measurable outcomes aligned to business priorities. Try-before-you-buy has become mission-critical for both buyers and sellers.

Read More