AI News Summary — Week 12.1, 2026

This week's AI landscape is dominated by dramatic shifts: xAI's internal crisis, Anthropic's legal battle with the U.S. government, Microsoft's new Claude-powered agent, and Hollywood's continued embrace of AI tools. Here's what's happening.

Elon Musk Admits xAI Needs "Full Rebuild" — 9 of 11 Co-Founders Gone

Elon Musk has acknowledge that xAI "was not built right" and is now "being rebuilt from the foundations up" The Rundown. In a candid X post, Musk admitted that Grok is "currently behind" — and the exodus of co-founders tells the story.

Key details:

• Zihang Dai and Guodong Zhang are the latest departures, leaving just two of the original 11 co-founders (Manuel Kroiss and Ross Nordeen) at xAI
• Zhang, who led Grok Code and reported directly to Musk, was reportedly blamed by Musk for Grok's coding shortfalls before departing
• xAI hired senior Cursor leaders Andrew Milich and Jason Ginsberg last week in an effort to catch up on coding capabilities
• The company is simultaneously preparing for one of the largest public listings in history

This represents a stunning turnaround for xAI, which just a year ago was riding high on $6 billion in funding and promises to challenge OpenAI and Anthropic. With Grok lagging behind the frontier and most of the founding team gone, Musk faces an uphill battle to deliver on those promises.

Anthropic Sues U.S. Government Over Pentagon Blacklist

Anthropic has filed two lawsuits against the Trump administration, challenging the Pentagon's "supply chain risk" label and a White House directive for all federal agencies to drop Claude The Rundown. The company calls the blacklist retaliation for its public advocacy on AI safety limits.

Key details:

• Anthropic is asking judges to throw out the blacklist label and block the government from forcing agencies to cut ties
• The company argues the "supply chain risk" label was designed to counter foreign adversary threats, not punish a U.S. company over policy disagreements
• Anthropic claims the Pentagon violated free speech rights by retaliating for publicly advocating AI safety limits on weapons and surveillance
• 30+ OpenAI and Google DeepMind employees signed a legal brief backing Anthropic, warning that the blacklisting threatens U.S. AI leadership

This is a remarkable turn of events: just months ago, Anthropic was one of the Pentagon's most prominent AI partners. Now it's suing the government in federal court. Whatever the outcome, the case could establish whether the government can penalize domestic AI companies for speaking out on safety issues — a precedent every AI lab will be watching closely.

Microsoft Launches Claude-Powered "Copilot Cowork"

Microsoft has introduced Copilot Cowork, a new Microsoft 365 feature built on Anthropic's Claude technology that runs tasks in the background across apps Microsoft. The feature operates in the cloud, pulling from emails, meetings, files, and chats across Microsoft 365 — a contrast to Claude Cowork's current desktop-only approach.

Key details:

• Microsoft built Cowork directly with Anthropic, using Claude Cowork's tech but wrapped in M365's enterprise security and compliance layers
• Users describe an outcome, and Cowork breaks it into steps, producing deliverables like decks, briefing docs, and workbooks across apps
• Cowork is available in a limited research preview, bundled with a new $99/user E7 tier that includes Copilot plus agent management and security tools
• The feature gives Copilot access to enterprise context across 450 million users' worth of emails, calendars, and files — something Claude Cowork can't easily match

"If you can't beat the thing that scared your investors, absorb it." Microsoft embedding Anthropic's agent tech inside M365's security boundaries represents a significant partnership in the enterprise AI space.

Netflix Acquires Ben Affleck's AI Filmmaking Startup

Netflix has acquired InterPositive, a stealth AI filmmaking startup that Ben Affleck started in 2022 Netflix. The deal brings all 16 staffers and Affleck himself aboard as senior adviser.

Key details:

• InterPositive's technology trains models on a production's own footage, then handles post-production work like relighting, swapping backgrounds, and fixing continuity errors
• Affleck emphasized the tech is "not generating video from nothing" — it learns from existing filmed shots and actors
• On a recent Joe Rogan Experience appearance, Affleck said he "can't stand" what AI writes, but sees the tech as a powerful tool for production workflows
• This is a rare acquisition for a streaming giant, and marks one of the most prominent Hollywood figures putting their reputation behind AI filmmaking tools

Hollywood has spent the AI boom either hiding the tech's use or railing against it. An Oscar-winning industry leader like Affleck backing the technology could shift sentiment significantly. The real upgrades, as Affleck addresses, come from production workflow aspects — not from generating video from nothing.

This roundup is brought to you by Conclio. Stay tuned for more AI news next week.