Skip to main content

PeakMetrics / Fact Checks

Technology & AI fact checks

Explore claims publishers have checked about technology & ai. Read their original ratings and follow the evidence.

1,815 published reviews in this topic

فکت‌نامهTechnology & AI

از تصاویر شهروندان در نمایشگر مترو مشهد به چه نکاتی پی می‌بریم؟

Publisher’s rating: هرچند جمهوری اسلامی قادر نیست از فناوری‌های هوشمند پردازش چهره بر اساس یادگیری عمیق برای جریمه افراد استفاده کند اما می‌توان در یگان‌هایی مانند قرارگاه ۲۱تیر، با تقسیم شهری و گماردن نیروی انسانی ناظر پشت دوربین‌ها، گزارش‌ها را با روش‌های دیگری در اختیار نیروی حاضر در محل قرار دهد.

Factcrescendo Sri LankaTechnology & AI

The Humanoid Waitress In A Chinese Restaurant !!(Video)

Publisher’s rating: Misleading

The claim concerns a viral video said to show a Chinese restaurant using a humanoid robot as a waitress. The review describes the footage as showing a woman serving food with robotic-like movements and says the video circulated on TikTok, X and Facebook. Its conclusion is that the restaurant did not employ a robot in the footage: the person shown was a dancer, and the restaurant owner was also featured in the clip. To investigate, the review says it consulted reporting by Business Insider and a video clarification from the South China Morning Post. It reports that both identified the person as a professional dancer rather than a humanoid robot. The review also says it found an interview with her by a Chinese television outlet during a dance competition, and that the footage showed her competing in dance events. These details identify the apparent robotic movements as a person’s performance, not evidence of a robot serving customers. The review notes that China has pursued humanoid-robot development and refers to the 2016 introduction of Jia Jia, described in the article as the country’s first humanoid robot. That broader technological context does not establish that the woman in this particular video was a robot; the review’s account of the specific footage identifies her as a dancer and the restaurant owner. Its stated conclusion is that the viral claim about a robot waitress is misleading.

dpaTechnology & AI

Facebook wird von jedem Nutzer eine Gebühr einziehen; mit einem eigenen Post kann man die Nutzungsbedingungen von Facebook aushebeln.

Publisher’s rating: Die Behauptung, Facebook werde von jedem Nutzer eine Gebühr einziehen, ist falsch. Auch die Behauptung, man könne mit einem eigenen Post die Nutzungsbedingungen von Facebook aushebeln, entspricht nicht den Tatsachen.

The dpa review examines a Facebook chain letter claiming that users would soon be charged a monthly fee and could stop Facebook using their pictures by copying and posting an opt-out declaration. The review says versions of this message circulated in Luxembourg and quotes the letter’s specific claim that Facebook would charge all users from Monday, while a copied post would supposedly let them opt out. Those are the same claims made in the check. The review gives a concrete reason the threatened bank withdrawal is implausible: Facebook generally did not have the bank details of ordinary users who were not advertising customers. It also distinguishes the rumor from a real paid option. Meta announced in a press release on October 30, 2023, that it would begin offering an ad-free version in November to customers in Germany, the European Economic Area and Switzerland, with a starting price of €9.99 per month. The review says users could still use Facebook with advertising for free. Thus, the paid service did not mean every user would automatically be charged. On the proposed opt-out post, the review cites the German consumer advice centre’s explanation that posting such a message could not change Facebook’s terms or resolve data-protection issues. Facebook’s terms allowed it to use published material for its own purposes, including displaying a photo to others on the network; uploading a photo did not, however, transfer its copyright. The review separately notes the centre’s criticism that aspects of Meta’s new terms concerning data collection and sharing with partners did not comply with data-protection rules. That qualification concerns the terms themselves, not a power for users to amend them by posting a chain message.

Folha de S.PauloTechnology & AI

Vídeo em que um homem afirma que uma “decisão da Anatel pode acabar com as lojas da Vivo, Tim e Claro no Brasil”. Em sua fala, ele utiliza um texto do site Terra Brasil Notícias concluindo que a mudança deixaria os brasileiros isolados do mundo e que ninguém estaria fazendo algo para evitar isso. O homem também faz menção, em tom conspiratório, ao decreto de Garantia da Lei e da Ordem (GLO) assinado no início do mês pelo presidente Luiz Inácio Lula da Silva (PT).

Publisher’s rating: Enganoso

The review examines a video in which a man says Brazil could be isolated from the rest of the world because of an Anatel decision that, as presented in the video, would sharply reduce telephone operators’ physical stores. The investigation identifies the underlying change as an amendment to the General Regulation on Consumer Rights for Telecommunications Services—not a change to the networks that carry communications. Under the earlier rule, operators had to maintain at least one in-person service store in each region with 100,000 or more inhabitants, with an additional unit for every 400,000 inhabitants. The revised rule ends that requirement: in-person service becomes optional, subject to accessibility rules, and store closures require authorization from the group implementing the changes. The changes were due to take effect in September 2024. The review reports that Anatel projected physical stores could fall from 2,800 to 789. It treats this as a significant reduction, but explains why that figure does not establish the video’s claim: the regulatory change does not alter operators’ transmission systems, and a decline in stores alone does not show that Brazil would be cut off from the rest of the world. The supplied review excerpt does not provide details about the video’s reference to a GLO decree.

dpaTechnology & AI

L'UE vient de voter en douce la mise en place de l'identité numérique. Ceux qui la refuseront seront exclus de la société.

Publisher’s rating: La dernière mouture du texte sur laquelle le Conseil et le Parlement se sont mis d'accord convient que le portefeuille d'identité numérique sera utilisé sur une base strictement volontaire. Les députés ont obtenu des garanties contre les discrimination envers ceux qui refuseraient de l'adopter.

The claim says the EU had quietly voted to impose digital identity and that people who refused would be shut out of society. The review examines the EU digital identity wallet, a proposed system for holding official documents on a smartphone and using them to identify oneself and access services. It reports that the European Parliament and Council reached agreement on the legal framework on 8 November 2023, but that the text still needed formal approval before becoming law. The review points to the negotiated text’s stated protections: use of the wallet would be voluntary, and people choosing not to use it should not face discrimination. It cites a European Parliament statement describing protections for citizens who do not use the wallet, and an explanation by Epicenter Works, an NGO collective member, that a person declining to use it in a given situation could not be charged more or denied access to a service or good. These provisions directly address the claim that refusal would mean exclusion. The review also notes qualifications rather than treating the agreement as a guarantee that every concern had been resolved. At the time covered, the agreement was not yet law; formal approval and a plenary vote were still ahead. It further reports that privacy and safeguards remained concerns for some experts and NGOs, including whether governments could learn too much about users’ activities. Those concerns are distinct from the claim that the wallet would be compulsory or that refusal would automatically exclude people.