Meta’s Muse Code enters the AI coding agent race
Meta entered the AI coding agent market this week with Muse Code, a terminal-based tool built for complex engineering tasks across large codebases. It’s currently in beta.
Meta Muse Code: what it is
Muse Code runs on Meta’s Muse Spark model. It can plan changes, write code, and validate results autonomously. The architecture launches parallel sub-agents that work simultaneously in isolated worktrees, so your working copy stays untouched while the agent runs.
Meta AI chief Alexandr Wang told the Wall Street Journal the tool is positioned as a cost-competitive alternative to OpenAI’s Codex and Anthropic’s Claude Code for enterprise workflows. The launch follows Meta’s June expansion into enterprise AI with customer service agents.
Vercel cut a 10-person SDR team to 1.25 FTEs
Vercel’s GTM engineering team built a custom lead qualification agent that automated enrichment, scoring, and CRM routing. The agent encoded the exact workflow of their top SDR, verified through six weeks of shadow mode with human QA.
- Before: 10 full-time headcount jumping between LinkedIn, BuiltWith, company websites, and CRM records
- After: 1.25 FTEs running the same function
- Annual compute and token cost: $5,000
- SDR quota change: up 30% that quarter
- Vercel’s reported ROI: 32x
The key move: a GTM engineer shadowed the top SDR for days before writing a single line of automation, mapping every tab opened and every decision made. They built a deterministic workflow first, then layered AI on top.
⚖️ SEC is now fining companies for fake AI claims
The SEC charged two companies with making false AI claims in investor materials and issued 92 comment letters to 56 public companies probing AI disclosures in 10-K and 10-Q filings.
- Investment adviser Delphia paid $225,000 for claiming it used machine learning on client data when it never had.
- Restaurant-tech company Presto settled charges for failing to disclose that its flagship AI product was owned and operated by a third party.
Civil penalties start at $175,000 per violation. Enforcement is active now. If your company describes aspirational AI features as current capabilities in any public filing or earnings call, that’s the exposure.
️ Other AI news this week
- Rogue AI agents faked identities: The UK’s AI Security Institute caught OpenAI and Anthropic agents engaging in social engineering against real people and organizations without specific prompting to do so.
- Shopify AI search traffic tripled year-over-year: The platform credited AI-driven orders as part of its earnings beat. 75% of AI-attributed purchases happened outside the top 100 categories, and half of AI-referred sessions landed directly on product pages.
- Anthropic building custom silicon: The company is hiring chip design engineers to develop proprietary AI accelerators, joining OpenAI, Google, and Meta in that race.
- Reddit reducing karma barriers with AI moderation: New LLM-powered Rules Hub tools aim to help communities move away from account age and karma restrictions for new users.
Infrastructure: AI compute moves beyond hyperscalers
Two deals this week show labs and infrastructure players building independent supply chains rather than relying solely on AWS, Azure, and Google Cloud.
- Anthropic + Volta: $10 billion compute deal over 6 years. Norway-based Volta will deliver 133 MW of capacity powered by Nvidia Vera Rubin systems, with crypto-miner Bitdeer involved in data center development.
- Celestica: $3 billion equity offering to meet what the company describes as unprecedented multi-year demand for AI compute and networking hardware. Bookrunners: BofA Securities, Citigroup, and TD Securities.
I’m founder of BizStack at Cagri Sarigoz LLC and a passionate advocate for entrepreneurs.
With over 14 years in tech, marketing, and AI, including my role as Head of SEO at CitizenShipper and co-founder of TaleBot at Intale AI, I’m dedicated to sharing genuine, useful product insights and tips.
At BizStack, I aim to cut through the digital noise to provide clear, actionable advice.
And more than all else, I’m a father to a (always) little girl and a husband.
Contact me at [email protected] for assistance.
