Good morning. It's Tuesday, March 10, and we're covering looming AI job shock, Anthropic’s showdown with the Pentagon, humanoid robots learning to clean homes, and more.

New here? Keep up with the future of tech, sign up here. If you enjoyed this email, share it with a friend.

Autonomous Living Room Cleanup
Helix 02, a humanoid robot controlled by a single neural system directly from visual input, now demonstrates autonomous living room cleanup. The system integrates walking, manipulation, and tool use to tidy cluttered spaces. It handles tasks like wiping surfaces, moving objects, and navigating tight areas. New skills were learned from additional data without changing algorithms. The demo highlights progress toward general-purpose household robots.

Anthropic Sues Trump Administration
Anthropic filed two federal lawsuits after the Pentagon labeled it a “supply chain risk” and ended a $200 million contract. The dispute followed talks where officials sought unrestricted military use of Claude, which Anthropic declined. President Trump ordered agencies to stop using the company’s products on February 27. Anthropic says the move punishes protected speech, while the Pentagon cites operational needs and shifted to OpenAI.

OpenAI Delays ChatGPT Adult Mode
OpenAI has delayed ChatGPT’s planned “adult mode” to focus on improving performance and product features. The feature, proposed in October 2025, would allow adult content for verified users. The company is prioritizing intelligence, personalization, and user experience for its 900 million users while rolling out age-prediction tools to protect minors. A senior hardware executive also resigned over concerns about the Pentagon AI contract.

OpenClaw AI Framework Surges Past 200K GitHub Stars
OpenClaw, an open-source AI agent framework launched in November 2025, has surpassed 200,000 GitHub stars amid rapid adoption in China. The platform lets users run AI agents across messaging apps to perform tasks like coding and web searches. Recent updates added memory fixes, security patches, and persistent chat. Growth may accelerate further following newly proposed Chinese government subsidies and tech integrations.

The Recap: Former U.S. Commerce Secretary Gina Raimondo argued in a New York Times opinion essay that artificial intelligence could trigger mass job displacement in the United States unless policymakers and businesses quickly build a new workforce transition system. Drawing on her experience implementing the CHIPS and Science Act and working with companies like TSMC, Raimondo says current education and training systems move far too slowly for the pace of AI-driven change. She proposes a public-private “grand bargain” in which employers are held responsible for defining essential skills and creating direct pathways into new jobs.

Highlights:

Forward Future Takeaways:
Raimondo’s argument reflects a growing concern among policymakers that AI’s labor impact could arrive faster than governments can respond. Her proposal shifts responsibility partly onto companies deploying AI, asking them to provide data and help design training pipelines tied to real jobs. → Read the full article here. (Paywall)

The Recap: Anthropic has filed two federal lawsuits against the Trump administration after being labeled a "supply chain risk" and barred from military contracts. The designation came after Anthropic asked for assurances its technology wouldn't be used for fully autonomous weapons or mass domestic surveillance. Legal experts say the company has a strong case. The law behind the designation is typically reserved for foreign firms suspected of espionage or sabotage, not American companies.

Highlights:

Forward Future Takeaways:
The clash highlights a growing tension between AI safety commitments and national security priorities as governments increasingly rely on commercial AI providers. If Anthropic prevails, the case could set an important precedent for how far the U.S. government can go in restricting domestic AI companies over defense concerns. → Read the full article here. (Paywall)

Early images of the Tesla Cybercab body-in-white suggest Tesla has created one of its simplest vehicle structures yet, reportedly using about 60% fewer parts than the Tesla Model Y. The design appears to rely on two large gigacastings—front and rear—combined with stamped panels and minimal joining components. Notably, the rear underbody casting uses flat lateral walls without integrated wheel arches, simplifying die design and eliminating the need for lateral slide cores.

Because the robotaxi has no steering column or driver controls, engineers were able to redesign the structural architecture from scratch. The approach aligns with Tesla’s planned “unboxed” manufacturing system, where major vehicle sections are built in parallel and joined during final assembly. → Read the full article here.

Nscale Raises $2B With NVIDIA: UK AI infrastructure startup secures $2 billion Series C led by Aker ASA and 8090 Industries, valuing it at $14.6 billion. Funding expands global data centers and cloud compute.

Microsoft Adds Anthropic to Copilot: Microsoft launches Copilot Cowork powered by Anthropic’s Claude models, expanding beyond OpenAI. The AI agent tool automates complex workplace tasks.

Alphabet Awards Pichai $692M Package: Google CEO Sundar Pichai receives a three-year compensation plan worth up to $692 million, largely performance-based and tied to stock incentives linked to Waymo and drone unit Wing.

Apple Eyes New ‘Ultra’ Devices: Apple is reportedly preparing premium products including a ~$2,000 foldable iPhone, an all-new touchscreen MacBook Ultra, and next-gen AirPods with cameras for visual AI context.

Anthropic Clashes with Pentagon Over AI Use: The AI firm refused U.S. military requests to allow Claude in mass surveillance and autonomous weapons, triggering a Pentagon “supply-chain risk” designation.

Thanks for reading. See you next time!

— Matthew Berman, Nick Wentz & the Forward Future Team