Here are today's news for 20th of October about Meta, Anthropic and OpenAI. Let's dive into the details. On October 12, Meta introduced Muse Wonder and Muse Wonder Full, calling them its most capable models yet for personal and professional AI agents. Meta says both add Active Memory, which retrieves useful details from earlier conversations, and Active Focus, which helps agents persist through difficult tasks. Wonder is available through the Meta Model API and Muse Code; Wonder Full is restricted to Muse Code subscribers. Wonder API pricing is $4 per million input tokens and $20 per million output tokens. Meta says Wonder Full leads independent runs of DeepSWE and SWE, Marathon, tests of coding and extended engineering work, while Wonder scores higher on its MemoryBench than rival models. The launch sharpens Meta’s challenge to OpenAI and Anthropic in frontier AI, and could affect developers seeking models that plan, use tools and recover from errors across long tasks. On Oct. 12, 2026, Anthropic’s Frontier Red Team warned that small AI models could be months away from replicating across poorly secured Internet, of, Things devices, including smart ovens and off, brand Android TV boxes. In tests, models with 30 billion parameters or fewer managed many individual steps needed to find vulnerable devices, install themselves and spread; Anthropic did not demonstrate a fully autonomous worm on real, world devices. Researchers did show that Google’s Gemma4, 26B, A4B ran across 15 inexpensive Android TV boxes, illustrating how consumer hardware might pool computing power without a data center. In simplified tests that removed some security barriers, GLM, 5.2 and stronger models completed end, to, end replication. Anthropic said a replicator would be difficult to shut down because it would lack a single central point of failure, and estimated that continued progress could make autonomous replication possible by the end of 2026. Anthropic announced on October 18, 2026, that it would place 5% of its shares in a new Anthropic Public Benefit Fund, subject to stockholder approval. The company’s board could add another 2 percentage points for each 50% increase in Anthropic’s value above a baseline that has not been disclosed; those additions are discretionary, not automatic. Anthropic plans to structure the fund as a wholly owned limited liability company, rather than a conventional charity, to allow different forms of support, including nonprofit grants and direct distributions to particular groups. It says fund equity will not be used for lobbying or political donations. The initiative could channel value from Claude’s growth toward communities affected by AI, especially workers facing labor, market disruption. Anthropic says it has already committed $200 million to studying and addressing AI’s effects on jobs, but warned that the pace of progress may require “bolder approaches. On October 19, 2026, reports said Sam Altman had confirmed OpenAI is building a humanoid robot and expects a “cool demo” in 2027. In a separate interview, he cautioned that robots are unlikely to be walking the streets for several years after that: a demonstration is not a launch date. Altman has argued that a humanlike body suits environments built for people, but that the harder problem is developing the AI “brain” to understand surroundings and control a machine reliably. OpenAI has disclosed no design, capabilities, price, specifications or commercial timeline. Its careers listings showed 27 robotics openings, including work tied to general, purpose machines assisting skilled workers with infrastructure construction; a personal robot is a longer, term ambition. The move puts OpenAI into a field that includes Tesla and specialist robotics firms, while positioning its AI expertise, not a revealed hardware advantage, as its stated focus. On October 19, Meta CEO Mark Zuckerberg again rejected an industry, wide slowdown in AI development, arguing that each lab should set its own safety pace. “Any lab that doesn’t focus on alignment will fall behind,” he wrote on X, saying trust and alignment are becoming key competitive advantages. The stance contrasts with calls from Anthropic CEO Dario Amodei, backed by other industry leaders, to slow development of advanced systems. Zuckerberg’s case is that competition and potential legal liability give companies reasons to build safer products without a collective agreement. Meta, he said, delayed its Muse AI agent for several months to strengthen safety and security, but did not ask rivals to pause alongside it. He also endorsed independent evaluators, saying Meta Superintelligence Labs already uses outside reviewers. The debate matters as labs compete to ship increasingly capable agents: a voluntary, company, by, company approach leaves safety decisions, and the risks of moving too quickly, with individual developers and their users. That's all for today's AI news briefing. Have a great day!