Guardian Labs' Guardião is a Portuguese AI startup that blocks phone scams before the call rings or the message is seen. The AI transcribes calls in real time, cross references scam databases, and detects manipulation tactics—all locally on the phone.
Research answer

Create a landscape editorial hero image for this Studio Global article: What is Guardian Labs' Guardião, how does its on-device AI stop phone scams before the call rings or message appears, what types of scams do. Article summary: Guardian Labs' **Guardião** is a Portuguese AI startup that blocks phone scams before the call rings or the message is seen. It was founded by Rita Barbosa after her grandmother lost €4,000 to a water-filter scam [1]. Th. Topic tags: general, general web, user generated. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fa
Guardian Labs' Guardião blocks phone scams before the call rings or the message is seen. No pop-ups, no alerts, no decisions required from the user. The app is designed primarily to protect elderly and vulnerable populations, and it runs entirely on-device—meaning no call audio, message text, or personal data ever leaves the phone .
Founder Rita Barbosa started building Guardião after her grandmother lost €4,000 to a water-filter scam . The experience drove home a painful reality: traditional spam-blocking tools often leave users—especially elderly ones—to spot and avoid scams on their own. Guardião flips that approach: instead of teaching people to distrust calls, the AI distrusts on their behalf
.
Guardião acts as an AI interception layer between an incoming call or SMS and the recipient . All AI models run locally on the phone, which means no call audio, message text, or personal data is sent to a remote server
.
The system checks known scam information and analyses content for tactics of manipulation . The user never interacts with the blocking layer at all.
Guardião is currently trained specifically on Portuguese scam patterns and targets multiple scam categories :
The system is currently country-specific (Portugal) by design, though expansion to other markets is planned .
The company has not published specific accuracy metrics, false-positive rates, or false-negative rates. However, the product's core design philosophy reveals its strategic tilt: the system is built to block aggressively (minimise false negatives) for elderly users who are unlikely to second-guess a scam . By intercepting before the user sees or hears anything, Guardião accepts that some legitimate calls may be blocked as the trade-off for near-total scam prevention. The company acknowledges this is a deliberate architectural choice for its target demographic
.
Guardião is currently in pre-seed / early-stage. The founder has stated that every euro from a recent prize (Google Gemini API Developer Competition) is going into development . Specific monetisation details—such as subscription fees, carrier partnerships, or B2B licensing—have not been publicly disclosed
.
As of August 2026, Guardian Labs has established several notable partnerships :
The company has also recently partnered with Subvisual, a Portuguese engineering firm, and Invisible Lab to accelerate product development .
Guardian Labs plans to expand beyond Portugal into other markets, with the on-device AI architecture designed to be trained on new regional scam patterns without sending data to the cloud . The company hired a Head of Operations in March 2026 and is actively seeking pre-seed funding to scale
. The longer-term vision is to make the phone "a safe place again" for elderly populations across multiple countries
.
Studio Global AI
This page includes a source-backed answer you can continue inside Studio Global.
Guardian Labs' Guardião is a Portuguese AI startup that blocks phone scams before the call rings or the message is seen.
Guardian Labs' Guardião is a Portuguese AI startup that blocks phone scams before the call rings or the message is seen. The AI transcribes calls in real time, cross references scam databases, and detects manipulation tactics—all locally on the phone.
Guardião targets impersonation scams, financial fraud, and phishing SMS, currently trained on Portuguese scam patterns.