Happy Wednesday! Last week OpenAI and Anthropic launched rival enterprise AI ventures while Microsoft debuted an AI legal agent in Word, Google pushed its AI control plane strategy, and Meta acquired a robotics startup for humanoid development.

AI ENTERPRISE PARTNERSHIP

Anthropic, OpenAI Launch Rival Enterprise AI Ventures

Anthropic and OpenAI have each announced joint ventures with major Wall Street firms to accelerate enterprise AI adoption. Anthropic is partnering with Blackstone, Goldman Sachs, and others in a $1.5 billion venture, while OpenAI's Deployment Company, led by COO Brad Lightcap, plans to inject up to $10 billion with partners including Bain Capital and SoftBank. Both ventures will embed AI engineers within client businesses to redesign workflows, initially targeting portfolio companies of their respective backers. Analysts note the efforts resemble traditional consulting firms, reflecting growing recognition that deploying AI in enterprises requires significant human expertise alongside the technology itself.

AI DEFENSE CONTRACTS

Pentagon Signs Classified AI Deals, Drops Anthropic

The Pentagon has secured classified AI agreements with OpenAI, Google, Microsoft, Amazon, Nvidia, xAI, and startup Reflection, granting the Defense Department access to their tools in sensitive settings. Notably absent from the new agreements is Anthropic, a previously trusted partner for classified work. The Defense Department dropped the AI company after designating it a supply-chain risk.

NEW GENERATIVE AI LAUNCH

xAI Launches Grok 4.3 With Aggressive Pricing

xAI has released Grok 4.3, a new large language model, alongside a voice cloning suite. The model shows significant benchmark improvements over its predecessor Grok 4.2 but still trails OpenAI and Anthropic's latest offerings, according to independent evaluator Artificial Analysis. The launch comes amid turbulence at xAI, including the departure of all 10 original co-founders and dozens of researchers. The company is positioning Grok 4.3 competitively on price as it battles rivals including OpenAI, Anthropic, Google, and several Chinese AI firms.

AI STARTUP ACQUISITION

Meta Buys Robot AI Startup for Humanoid Push

Meta Platforms has acquired Assured Robot Intelligence, an AI software startup for robots, to advance its humanoid robotics ambitions. Financial terms were not disclosed. The San Diego-based startup was co-founded by AI researchers Lerrel Pinto and Xiaolong Wang, whose work includes a humanoid robot control system using VR headsets. The acquired team will join Meta Superintelligence Labs, which oversees the company's AI and robotics efforts. Meta reportedly aims not to build its own humanoid robot but to supply software and components to other market players, similar to Qualcomm's role in the mobile industry.

NEW AI AGENT LAUNCH

OpenAI Rushes AI Agent Phone for 2027

OpenAI appears to be fast-tracking its first AI agent phone, with mass production targeted as early as the first half of 2027, according to TF International Securities analyst Ming-Chi Kuo. The accelerated timeline may be driven by a desire to support a year-end IPO narrative, Kuo suggested. The development signals OpenAI's ambitions to expand beyond software into consumer hardware devices.

AI LABOR UNION DISPUTE

DeepMind Workers Unionize Over Military AI Contracts

Google DeepMind staff at its London headquarters have voted to unionize, with 98 percent of CWU members supporting the move. Employees are demanding the company stop its AI technology from being used by Israel and the US military, with one worker stating the work is "helping make genocide cheaper, faster, and more efficient." The bid would represent over 1,000 London-based staff. Google has 10 working days to voluntarily recognize the unions before legal action begins. Workers are also considering research strikes and protests, following Google's recent Pentagon AI deals and the 2024 firing of over 50 employees who protested military contracts.

AI ENTERPRISE FORECAST

AI to Make Half of Business Decisions by 2030

A new IBM poll reveals that CEOs expect artificial intelligence to autonomously handle 48% of operational decisions by 2030. The shift is anticipated primarily in areas governed by clear, established rules and procedures. The findings highlight growing executive confidence in AI's role in business operations, signaling a major transformation in how companies will be managed within the next five years.

NEW GENERATIVE AI LAUNCH

Microsoft Launches AI Legal Agent Inside Word

Microsoft is introducing a Legal Agent inside Word, designed specifically for legal teams to handle contract reviews, document edits, and negotiation history. The agent follows structured workflows based on real legal practice, reviewing contracts clause by clause and flagging risks and obligations. The tool is currently rolling out to US members of Microsoft's Frontier program. It draws on talent acquired from Robin AI, a failed startup that built AI-powered contract review technology.

GENERATIVE AI LEGAL DISPUTE

Musk Admits xAI Trained Grok on OpenAI Models

Elon Musk testified in a California federal court that xAI used distillation techniques on OpenAI models to train Grok, calling it a general industry practice. When pressed for a direct answer, he confirmed it was partly true. The admission came during his ongoing lawsuit against OpenAI and CEO Sam Altman over the company's shift to a for-profit structure. Distillation involves systematically prompting AI models to replicate their capabilities at lower cost, threatening the competitive advantage of frontier labs. Musk also ranked AI leaders during testimony, placing Anthropic first, followed by OpenAI and Google, while describing xAI as a much smaller operation.

GENERATIVE AI CLOUD INSIGHTS

Google Bets on Control Plane to Win AI Race

Google Cloud Next 2026 revealed that the real AI battleground is infrastructure, not models. Analyst John Furrier argues that whoever controls the agentic control plane — the layer connecting data and systems across the enterprise — wins the AI era. Key insights include the critical role of contextual data in agentic AI, with experts from Google Cloud and OpenText stressing that models need the right data at the right time, not just raw storage. Enterprises are also restructuring teams and decision-making around AI-driven token economies, signaling a fundamental corporate reset.

Upcoming Data & AI Conferences

Google I/O 2026

Tue, May 19 - Wed, May 20, 2026 | Mountain View, CA

Curated with AI by EMC2 AI

Brought to you by Estevan McCalley

Share this issue

Follow Estevan

Like this? Subscribe for the weekly AI roundup.