Realtime AI News
Meta Brings Its Muse AI Agent to Mexico, Its First Latin American Market
Meta has launched its personal AI agent Muse in Mexico, the first Latin American market for the product, with Spanish-language access for users 18 and older via iOS, Android and the web. The Mexican version connects to local services including Liverpool, Rappi, Cinépolis and Kavak, with Meta saying those connectors will roll out gradually.

Meta has launched its personal AI agent Muse in Mexico, according to Mexico News Daily, making the country the first Latin American market for the product. Muse had previously been available in the United States, so the Mexican rollout marks the company's first attempt to take the agent beyond an English-language market.
Access is limited to users 18 and older and runs in Spanish. Mexicans can use Muse through the iOS and Android apps and the muse.ai website, or interact with it directly inside WhatsApp. Meta has also said Muse will later work with its smart glasses.
Muse is positioned as more than a chat box. Per Meta's description, it can keep working after the user closes the app: following up on pending tasks, helping manage projects, and turning long-term goals into concrete action plans. That ability to act on the user's behalf, rather than just answer questions, is the core distinction from a conventional assistant.
For the agent to actually get things done, the Mexican version plugs into a local services ecosystem. The first partners span retail, credit and lending, travel, car rental and buying, entertainment and food delivery, and include Liverpool, Coppel, Rappi, Cinépolis, Despegar, Kavak and Kueski. Meta said explicitly that these connectors will be enabled progressively, with some features arriving after the initial launch.
On the general-purpose side, Muse can also connect to Gmail, Google Calendar and Instagram to handle email, schedule the day, or draw on content a user has saved. These integrations are opt-in and form the main channel through which the agent builds context about a person's life.
Security and privacy are what Meta emphasized most in the launch materials. According to the company, Muse runs inside a dedicated, protected computer with its own separate browser, and was designed from the outset with security and privacy as central constraints. Given that a personal agent needs access to email, calendars and potentially payments, whether that isolation holds up in practice will determine how much access users are willing to hand over.
On pricing, Muse is free to use, with optional subscription tiers that unlock additional features. That free-entry, paid-depth structure mirrors how Meta has long run consumer products, and it places the monetization burden on feature depth rather than on access.
Strategically, Mexico is more than an added market for Meta: it is the first full test of wiring agentic task execution into a local commercial ecosystem. Latin America is a key region for Meta's user growth, and local retail, mobility and delivery platforms are exactly where an agent can most easily prove its value. If the Mexican connectors work, the model becomes replicable across Spanish-speaking markets. What to watch next is the pace of connector enablement, actual subscription conversion, and whether regulatory and privacy scrutiny slows the expansion.
Why it matters
The Mexico launch moves Muse from a single US market into Latin America and ties it to local retail, mobility and delivery platforms for the first time, making it the template for how far Meta's agent can scale internationally.
Nearby Updates
All11/07, 06:53
OpenAI Has Raised $11B of a New Round at a $1.25T Valuation
OpenAI has reportedly raised $11 billion so far in a new funding round that values the company at $1.25 trillion, according to Bloomberg reporting cited by TechCrunch. The round has not closed and could total roughly $20 billion within weeks — well short of the $30 billion at a $1.4 trillion valuation OpenAI was said to be targeting in late September.
11/07, 04:41
Chinese open models near OpenAI's gated frontier system on autonomous hacking, at a tenth of the cost
A Dealroom report published November 6 says an evaluation of autonomous cyber capabilities found Chinese open-weight models approaching the frontier model OpenAI offers only through a restricted security program, at roughly a tenth of the cost per attempt. The same evaluation warns that once weights are public, no provider can monitor or stop how a model is used.
11/07, 03:54
Benzinga: Anthropic Holds About 60% of Business AI Spend as OpenAI Falls Behind
Benzinga reports that Anthropic accounts for roughly 60% of business AI spending, with OpenAI described as falling behind. The split suggests enterprise buyers are concentrating budgets on a single vendor, a shift with direct consequences for pricing, contracts and lock-in risk.
11/07, 02:44
Hundreds of Artex AI hacking tools found exposed on the open web
Security researchers found hundreds of servers hosting the penetration testing tool Artex openly reachable on the internet. Anyone can access the hacking tool without authorization.