AI Credit Scoring Certification Fee Scam: Impersonating Teddy Open Finance to Collect 200 BRL Advance Payments in Brazil

Who Gets Targeted

Victims are primarily low-income populations, micro-business owners, individuals with poor credit histories, or those unable to obtain traditional bank credit in Brazil, as well as some Chinese cross-border practitioners attempting digital credit in Brazil. They hold a technological worship of 'AI credit scoring,' mistakenly believing AI can bypass bank risk control to secure loans, while fearing in-person bank audits would expose income issues. Consequently, they prefer completing applications via WhatsApp chats. When in urgent need of money, scammers simply dangle scripts such as 'score approved, certification fee refunded upon disbursement,' exploiting their anxiety and trust in authoritative brands to harvest funds.

骗局怎么运作

• Step 1: Brand Impersonation Ad Traffic Generation. Scammers purchase keywords like 'Teddy AI Scoring' and 'No-Credit-Check Loans' on Instagram, Facebook, and search engine ad spaces, stealing official website screenshots and logos of Teddy Open Finance to create ads claiming 'AI credit scoring system spring expansion, limited slots available,' inducing users to click, direct message, and fill in their name, phone number, CPF (Brazilian tax ID), and desired loan amount.
• Step 2: Forged Scoring Report to Establish Credibility. A fake customer service agent sends an 'AI credit scoring report' featuring Teddy's logo via WhatsApp, showing a comprehensive score above 80 and a pre-approved loan limit, along with a formal-looking electronic contract link. The script emphasizes that 'certification is solely used to confirm your identity and will be automatically refunded via the original payment method after disbursement,' making users think this is standard procedure for a regular financial institution.
• Step 3: Pix Prepayment to Lock in Funds. The agent requests a 200 BRL 'certification fee' and provides a personal bank account or Pix collection code. In Brazil, Pix transfers are credited instantly and cannot be reversed like credit cards; scammers exploit this exact mechanism, immediately transferring funds to multiple accounts upon receipt, causing victims to lose control the moment they pay.
• Step 4: Deepfake Identity Verification for Ceremonial Authenticity. Scammers require users to record a selfie video saying 'My name is XX, and I am voluntarily applying for a loan' while holding their ID card. They then use AI face-swapping tools to generate a video response featuring a 'reviewer wearing a Teddy uniform' with synchronized lip movements, lying that the manual review has passed. This both tricks the user into thinking they have entered formal risk control procedures and seizes the opportunity to collect facial biometrics and identity documents.
• Step 5: Secondary Harvesting and Disappearance. When users repeatedly follow up on loan disbursement, customer service continues to demand further transfers under pretexts such as 'system score invalidation,' 'central bank risk control requiring additional margin deposits,' or 'unfreezing quota requirements.' If the user refuses, scammers threaten that 'the contract is effective and failure to process will result in being blacklisted on credit bureaus,' subsequently logging out of WhatsApp accounts, deleting pages, and reselling victims' personal information and facial videos to the black market for synthetic identity fraud.

红旗信号（看到这些快跑）

• 🚩 The counterparty cannot provide a financial institution authorization number issued by the Central Bank of Brazil (Banco Central do Brasil), and their entity cannot be found in any official institutional list.
• 🚩 Demanding 'certification fees,' 'margin deposits,' or 'unfreezing fees' paid to personal accounts via Pix. Legitimate Brazilian lending institutions never require upfront payments before loan disbursement.
• 🚩 Promotional materials show obvious AI traces: distorted areas in promotional images, out-of-sync lip movements and audio tracks in customer service videos, template-based replies, and refusal to engage in real-time video calls.
• 🚩 Website domain registration duration is short, and the page lacks a real office address, CNPJ (Brazilian corporate tax ID), or executive information, featuring only WhatsApp and Telegram entry points.
• 🚩 The entire process heavily emphasizes 'no credit check, instant approval, internal channels,' while continuously manufacturing a sense of urgency with 'last slots today' and 'system closing soon'.

真实案例

• In August 2026, the Brazilian newspaper O Globo reported that fraud syndicates posted 'guaranteed loan' advertisements on social media using the names of major banks and credit brands, inducing consumers to pay 'processing fees' via WhatsApp beforehand. The implicated financial institutions subsequently issued statements denying any cooperation with intermediary agents, with involved amounts ranging from hundreds to thousands of BRL. (Source: [https://extra.globo.com/economia/noticia/2026/08/golpistas-utilizam-nomes-de-grandes-marcas-em-falsas-ofertas-de-emprestimos-saiba-como-identificar.ghtml](https://extra.globo.com/economia/noticia/2026/08/golpistas-utilizam-nomes-de-grandes-marcas-em-falsas-ofertas-de-emprestimos-saiba-como-identificar.ghtml))
• In June 2026, China Brazil Insight reported that generative AI drove a twofold increase in financial fraud reports in São Paulo, significantly increasing risk control pressure for some Chinese-funded cross-border financial institutions in Brazil. Cases cited in the report already included modus operandi where AI-synthesized customer service voices impersonated lending platforms to guide users into making Pix transfers.
• In the first half of 2026, Serasa Experian released financial fraud statistics showing a nearly 10% year-on-year increase in fraud attempts within the financial sector, focusing heavily on pre-loan fee collection and identity theft scenarios. The institution also pointed out that AI-generated documents and liveness videos are making traditional anti-fraud model identification significantly more difficult.
• In June 2023, Brazilian news portal G1 reported that a criminal syndicate specializing in fake loan scams targeting retirees and pension recipients was dismantled by police in Niterói, Rio de Janeiro. 11 suspects were arrested, and police discovered notes inside the syndicate's office detailing targets of up to 320,000 BRL in involved funds. (Source: [https://g1.globo.com/rj/rio-de-janeiro/noticia/2023/06/16/quadrilha-do-falso-emprestimo-classificava-vitimas-com-estrelinhas-tinha-metas-a-cumprir-e-olheiro-para-evitar-policiais.ghtml](https://g1.globo.com/rj/rio-de-janeiro/noticia/2023/06/16/quadrilha-do-falso-emprestimo-classificava-vitimas-com-estrelinhas-tinha-metas-a-cumprir-e-olheiro-para-evitar-policiais.ghtml))
• On August 28, 2026, the Brazilian Association of Software Companies (ABES), citing Serasa Experian data, pointed out that in the first half of 2026, Brazil's financial sector suffered a total of 1,747,373 identity fraud attempts, a year-on-year increase of 9.5%, with anti-fraud technology intercepting an estimated 100.4 million BRL in involved funds. (Source: [https://abes.org.br/setor-financeiro-enfrentou-alta-de-quase-10-nas-tentativas-de-fraude-no-1o-semestre-revela-serasa-experian/](https://abes.org.br/setor-financeiro-enfrentou-alta-de-quase-10-nas-tentativas-de-fraude-no-1o-semestre-revela-serasa-experian/))

Official Stance

• In July 2026, Serasa Experian released a semi-annual fraud alert, reminding consumers to remain vigilant against AI-generated loan advertisements and 'credit score' screenshots, noting that any fee collection prior to loan disbursement is highly suspicious.
• In August 2026, the economics section of Brazil's O Globo cited consumer protection agency recommendations, urging the public to verify credit institution brand authorization and stay alert against false loan ads impersonating major brands.
• Long-standing consumer warnings from the Central Bank of Brazil state that unauthorized institutions are prohibited from engaging in credit intermediation activities, regular lenders must not require consumers to pay 'certification fees' or 'unfreezing fees' prior to loan disbursement, and suspected entities can be reported through official central bank channels.

How to Protect Yourself

• ✅ Check Licenses First: Visit the 'Instituições Autorizadas' list on the official website of the Central Bank of Brazil to verify whether the counterparty has credit business authorization; if no number can be found, it can be deemed a fake institution, and contact should be terminated immediately.
• ✅ Refuse All Upfront Payments: Fees for legitimate Brazilian loans can only be deducted from the disbursed loan amount. Any 'certification fees' or 'margin deposits' paid via Pix or prepaid cards should be treated as fraud.
• ✅ Independent Verification: Call the customer service numbers published on the official websites of banks or financial institutions to cross-check, instead of using numbers sent via SMS and WhatsApp. At the same time, request the counterparty to send a formal contract via corporate email attached with their CNPJ.
• ✅ Seize the Pix Recovery Window After Being Scammed: Immediately contact your account-holding bank to apply for the Central Bank of Brazil's Special Return Mechanism (MED), save transfer receipts, chat histories, ad screenshots, and website links, report the case to the police, and synchronize with the central bank's consumer complaint channels.