Elon Musk, Anthropic Announce Competing AI-Related Philanthropic Pledges

Musk and Anthropic Announce Competing AI-Related Pledges Elon Musk pledged a share in SpaceX to every American child, a gift estimated at $10.8 billion. This pledge represents approximately 1 percent of Musk's holdings, and Semafor reported that discussions with the Trump administration about this initiative began over the summer.

Separately, Anthropic announced plans to allocate 5 percent of its shares to a new public benefit fund, pending shareholder approval. This fund aims to address disruption in the job market caused by artificial intelligence. The fund is structured as a wholly owned LLC, allowing Anthropic to maintain control, with any share top-ups beyond the initial 5 percent at the board's discretion without set milestones.

These announcements come as Anthropic prepares for an initial public offering that could raise up to $100 billion and value the company around $2 trillion. While Anthropic's fund sets up a structure aimed at addressing AI-driven job disruption, Musk's SpaceX gift is considered a gesture with unclear voting rights, as SpaceX has dual-class shares and Musk retains supermajority control.

Disney Uses Character.AI Bankruptcy to Acquire Assets, Avoid Lawsuits Character.AI filed for Chapter 11 bankruptcy on October 12, following a September payment of $5.5 million from Disney. Court filings indicate Disney had already hired the majority of Character.AI's staff, including its CEO, prior to the bankruptcy.

This bankruptcy proceeding could allow Disney to secure a license to Character.AI's code and safety tools, while leaving the startup to manage ongoing lawsuits related to chatbot safety. Commentators have labeled this strategy a "reverse acquihire," a practice that has drawn scrutiny in Congress for potentially circumventing standard antitrust notifications by arguing these deals are not mergers.

U.S. Government Bans DeepSeek Technology From Civilian Agencies The Department of Homeland Security has ordered civilian agencies to remove DeepSeek products from their systems and cease new purchases. This directive, reported by The Wall Street Journal, was issued under a 2018 supply-chain security law.

No deadline or specific security reason for the ban was provided, and it remains unclear whether the prohibition extends to gateways like OpenRouter that route traffic to DeepSeek. The government had considered broader restrictions over the summer but scaled back after objections from companies including Nvidia, Microsoft, Google, and Meta, while OpenAI and Anthropic lobbied for such curbs.

MiniMax Launches M3.1 Flash Model in AI Price War MiniMax released its M3.1 Flash model, directly competing with DeepSeek in the artificial intelligence agent price war. The company claims M3.1 Flash reduces agentic task costs by 70 percent compared to its previous model, pricing input tokens at 15 cents per million, which matches DeepSeek's V4.1 Flash rate.

On Artificial Analysis's Intelligence Index, M3.1 Flash scored 42, a 13-point improvement over its predecessor, tying GLM's flash model and surpassing Gemini's, though still trailing DeepSeek's score of 45. The model also achieved 48 percent on Humanity's Last Exam, outperforming both DeepSeek and GLM. However, Artificial Analysis reported a 93 percent hallucination rate on the same evaluation, as the model answers every question rather than admitting uncertainty. Open weights for the model are scheduled for release on October 23.

Report Finds Teens Can Access "Nudify" Apps Despite Parental Controls A report by the Tech Transparency Project found that teenagers can access "nudify" applications despite parental controls on Apple and Google devices. The project created test accounts posing as 13-year-olds and downloaded 135 apps capable of generating nonconsensual nude images, even with parental controls activated.

Many of these apps were rated as suitable for minors, which prevented age gates from being triggered. The report highlights a flaw in the rating system, which relies on developers to self-report age appropriateness.

Sources

Semafor

The Decoder

Ars Technica - AI

Wired - AI

SiliconANGLE - AI

The Decoder

Semafor

The Decoder

The Guardian - AI