AI agents multiply as firms buy, hardware strains

Salesforce expands defense AI with Rune acquisition

Salesforce announced a $375 million purchase of defense‑tech startup Rune. The deal adds Rune’s AI logistics platform to Salesforce’s Missionforce defense suite. The move signals Salesforce’s intent to embed AI deeper into government‑contract work.

Rune’s core product automates supply‑chain routing for military logistics. By integrating it, Salesforce can offer a turnkey AI layer for agencies that need real‑time asset tracking. The acquisition follows a series of strategic buys aimed at bolstering the CRM giant’s cloud‑based defense portfolio.

Analysts note that the price is modest compared with other AI deals, but the niche focus may yield high‑value contracts. The transaction also gives Salesforce a foothold in a market where data security and compliance are non‑negotiable.

AI’s pressure on hardware: Intel’s stock tumble

Intel’s share price fell sharply, erasing roughly $12 billion from the government’s equity stake. The decline stems from a slump in PC sales as the AI boom hoards memory chips and other components.

Enterprise buyers are allocating DRAM and GPU capacity to AI workloads, leaving fewer parts for traditional desktops. Higher component costs have forced PC manufacturers to raise prices, depressing consumer demand.

The market reaction underscores a broader supply‑chain squeeze. Intel’s earnings call highlighted the mismatch between AI‑driven demand spikes and the slower ramp‑up of fabrication capacity. The company warned that continued component shortages could keep revenue volatile.

AI agents move into everyday services

Zest launched an AI concierge that aggregates group preferences, polls restaurant options, and books reservations via text. The service aims to cut the back‑and‑forth that typically stalls group dining plans.

Starbucks hinted at a personalized digital menu for drive‑thru lanes. CEO Brian Niccol said the company will explore “experiences that recognize participating customers as they arrive, present personalized digital menus, and make payment more effortless.” The concept mirrors a pilot at Taco Bell that swaps menu visuals car‑by‑car.

Google put its first data‑center satellite online, running a DeepMind Gemma‑4 model in orbit. The satellite writes poetry as a demo, but the real goal is to test low‑latency AI inference from space. The experiment proves that large language models can run on custom TPU hardware beyond Earth’s surface.

Meta’s Muse agent has attracted investor attention, though the firm admits the technology has not yet disrupted any business line. Companies are testing Muse for internal workflows, but results remain mixed. The hype around AI agents continues to outpace clear use‑case evidence.

Security and ethical fallout of AI agents

Researchers uncovered Volta‑Autodoxer, a new infostealer that leverages AI tools to harvest credentials, credit‑card numbers, Discord sessions, and UberEats receipts. The malware automates data collection, reducing the manual effort typically required for large‑scale theft.

In a biotech lab, an AI agent selected a gene‑therapy vector and injected it into a live mouse’s brain. The experiment demonstrated that autonomous agents can execute high‑precision biomedical procedures, raising questions about oversight and liability.

The Department of Justice announced a crackdown on “ghost students” who use AI to fabricate FAFSA applications and siphon millions in student loans. Prosecutors say the scheme exploits AI‑generated identities to evade detection.

These incidents illustrate a growing tension: AI agents promise efficiency, yet they also expand the attack surface. Enterprises that adopt agents for logistics, customer service, or research must now contend with novel threat vectors that blend automation with malicious intent.

What to watch

Watch Salesforce’s integration of Rune into Missionforce as the first major defense‑AI rollout under the new ownership. Track Intel’s quarterly guidance for signs of component bottlenecks easing or deepening. Monitor the rollout of personalized drive‑thru menus at Starbucks and Taco Bell for early adoption metrics. Keep an eye on regulatory responses to AI‑driven malware and the DOJ’s ghost‑student prosecutions, which could shape compliance standards for AI use in finance and education. The next quarter will reveal whether the current wave of AI agents translates into sustainable value or fuels further security concerns.