Today in AI — 14 May 2026

Cerebras raised $5.55 billion in its IPO. Alibaba posted its first operating loss since Covid. Cisco hit record revenue while cutting nearly 4,000 jobs. The infrastructure bet is enormous, and today's question is who fills all that compute with products people actually use.

The capex bet

AI infrastructure spending is accelerating faster than the software that runs on it. Cerebras priced at $185 per share. Cisco's stock surged 17% on record revenue despite cutting nearly 4,000 jobs. Alibaba took an operating loss to chase 38% cloud growth.

• Cerebras prices IPO at $185 per share, raising $5.55 billion as shares debut on Nasdaq — CNBC
• Cisco cuts nearly 4,000 jobs in AI restructuring as stock surges 17% on record revenue — CNBC
• Alibaba posts first operating loss since Covid as AI spending surges, but cloud revenue jumps 38% — CNBC
• Forget Grok — Musk's real AI edge is infrastructure, not software — Bloomberg

Agents as the interface

AI is moving from sidebars and chatbots into the primary interaction layer. Baidu proposed Daily Active Agents as the new success metric. Amazon replaced Rufus with an agentic shopping assistant. Workday and Microsoft are embedding agents directly into enterprise platforms.

• Baidu declares the shift from model competition to agent competition, proposes Daily Active Agents metric — TechNode
• Amazon retires Rufus chatbot and launches Alexa for Shopping as an agentic commerce assistant — CNBC
• Workday plugs Sana AI agent into Microsoft 365 Copilot for HR and finance self-service — Workday
• Microsoft retires Copilot Mode in Edge, bakes AI features directly into the browser — Microsoft Edge Blog
• Anthropic's Cat Wu says Claude will soon anticipate your needs before you know what they are — TechCrunch

Cybersecurity's AI arms race

Palo Alto Networks warns AI-driven cyberattacks will become the norm within months. The ECB is urging eurozone banks to prepare, and Mistral is building a cybersecurity model for those that lack access to Anthropic's Mythos.

• Palo Alto Networks warns AI-driven cyberattacks will become the 'new norm' within months — CNBC
• ECB urges eurozone banks to urgently prepare for AI-assisted cyberattacks — NL Times
• Mistral developing cybersecurity AI model for European banks lacking Mythos access — Bloomberg

Vertical AI and governance

Anthropic launches Claude for Legal with 12 practice-area plugins and 20+ software connectors. OpenAI backs a global AI governance body led by the US and including China. And Musk's xAI requires SpaceX IPO advisers to buy Grok subscriptions.

• Anthropic launches Claude for Legal with 12 practice-area plugins and 20+ software connectors — TechCrunch
• OpenAI backs creation of global AI governance body led by the US and including China — Bloomberg
• Musk's xAI races to get Wall Street firms to use Grok, requires SpaceX IPO advisers to buy subscriptions — Bloomberg

The infrastructure is being built and the agents are being shipped. For builders, the uncomfortable question: when the primary interaction becomes "the AI does it for you," what's left to differentiate?

---