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

Boatos.orgTechnology & AI

TikTok anuncia que chegou ao fim e vai encerrar atividades em 30/06

Publisher’s rating: Falso

The claim under review was that TikTok had announced it would end all operations on June 30, 2023. The review says versions circulating on social media described TikTok as ending or ceasing operations on that date, and explains that the story arose from a real announcement by ByteDance, TikTok’s owner, but applied it to the wrong service. The service named in the announcement was Helo, a separate social network owned by ByteDance. The review reproduces a message sent to Helo partners stating that the app would end its services and operations starting June 30, 2023. It also quotes a ByteDance statement to the press saying it had decided to end Helo’s services and that the app would stop working on June 30. The review concludes that this Helo shutdown was misrepresented as the end of TikTok. It says ByteDance made no announcement that TikTok was ending its activities, and identifies the confusion between the two services as the basis for the claim. The evidence presented therefore supports a distinction between a real closure date for Helo and the unsupported assertion that TikTok would close on that date.

Fact Crescendo Sri LankaTechnology & AI

ඉලොන් මස්ක් රොබෝ භාර්යාවන් නිර්මාණය කළා ද?

Publisher’s rating: False

The review investigates social-media posts claiming that Elon Musk and Tesla were making “robot wives,” with an alleged September 2023 release and claims about charging, operation and price. It also examines images presented as showing Musk with the robots. The investigation says those posts and images—not an announced Tesla product—were the subject of its check. Its conclusion is that the robot-wife claims were false and that Musk and Tesla had made no official announcement of such a project. (S1: “Elon Musk සහ ඔහුගේ සමාගම Robot Wife කෙනෙක් සෑදීමේ අවසන් අදියරේ ඉන්නේ. රොබෝ භාර්යාවන්ගේ සැපයුම 2023 සැප්තැම්බර් මාසයේදී නිකුත් කිරීමට අපේක්ෂා කෙරේ.”; “එවන් ප්‍රකාශයක් ඉලොන් මස්ක්හෝ ටෙස්ලා සමාගමින් කිසිදු නිල තොරතුරක් දක්වා නොමැති බවයි.”) For the images, the review describes visual inspection: it points to blurred areas near fingertips and extra fingers as irregularities that led it to identify the pictures as AI-generated. It also reports that the images were associated with digital creators: it names Guerrero Art in connection with Musk-and-robot images, and says Diana Stark posted prompts and resulting images showing Musk with a robot in different poses. These details support the review’s conclusion that the circulated pictures were not evidence that Tesla had built robot wives. (S1: “එහි ඇඟිලි තුඩු අසල බොඳ වී තිබීම , වැඩිපුර ඇඟිලි පිහිටීම වැනි සියුම් විකෘතිතා දැක ගන්නට ලැබෙනවා.”; “Diana Stark නම් ඩිජිටල් නිර්මාණ සිදුකරන කාන්තාවක්, ට්විටර් සටහනක් තබමින්, මේ ආකාරයේ AI භාවිතයෙන් Elon Musk සමඟ රොබෝවෙකු විවිධ ඉරියව්වලින් සිටින ඡායාරූප නිර්මාණය කිරීමට යොදා ගත් ඛණ්ඩ ඇතුළත් කරමින්, ඒ මඟින් නිර්මාණය වූ ඡායාරූප ද පළකර තිබුණේ මෙලෙසින්.”) The review distinguishes the viral claim from Tesla’s humanoid-robot project. It says Tesla had recently shown a video of its Tesla Bots walking, picking up items and identifying objects, and describes improvements including motor-torque control and AI training for handling objects. It characterizes those as the bots’ demonstrated abilities—not the claimed functions, commercial availability or specifications of robot wives. (S1: “එම රොබෝ පිරිස මේ වන විට ක්‍රමානුකූලව ඇවිදීමට, අයිතම ලබා ගැනීමට සහ වස්තූන් හඳුනා ගැනීමට හැකියාව ඇති බව දැක ගන්නට ලැබේ.”; “වැඩි දියුණු කළ මෝටර් ව්‍යවර්ථ පාලනය, මිනිසා විසින් නිරීක්ෂණය කරන ලද චලනයන් සහ වස්තු හැසිරවීමේ හැකියාවන් මත පදනම් වූ කෘතිම බුද්ධි (AI) පුහුණුව ඇතුළත් වේ.”)

Factcrescendo Sri LankaTechnology & AI

Elon Musk’s Humanoid Robot Wives !!(Images)

Publisher’s rating: False

The review investigated social media images presented as showing Elon Musk with female humanoid robots who were supposedly his wives. The posts described the women as advanced robots created using artificial intelligence, and made claims about their personalities and features. The review’s conclusion was that the images did not show real scenes of Musk with robots: it identified them as AI-generated. To investigate, the publisher says it searched mainstream media for reporting about Musk’s alleged robot wives and checked his official social media accounts, finding no such reports or posts. It also examined the images for visual irregularities, including abnormal fingers, unusually smooth appearances and blurry backgrounds, which it described as characteristics of AI-generated imagery. The review then used reverse-image searches to trace the images to digital creators known for AI-generated work. The review names art_is_2_inspire on Instagram, whose Facebook page is called Guerrero Art, and Diana Stark as creators connected to related images. It reports finding two images related to the claim on the art_is_2_inspire Instagram page and other related images and prompts in a Diana Stark Twitter post. Separately, it notes that Tesla had a humanoid robot program, Tesla Bot, but says it was not as advanced as the robots portrayed in the viral posts. The evidence presented therefore concerns the authenticity of the specific viral images—not the existence of humanoid robots generally.

Fact Crescendo Sri LankaTechnology & AI

කෘතීම බුද්ධියෙන් නිපදවන ඡායාරූප හදුනාගන්නේ කෙසේද? ලෝකයේ අවසාන යෝධයාගේ සේයාරුව ඇසුරින් පහදමු.

Publisher’s rating: False

The review examined an image circulated with the caption “the world’s last giant.” It says that some versions also described the image as dating from 1901 or showed the “last Neanderthal.” The investigation focused on whether the image was a genuine old photograph and used this example to discuss how AI-generated images may be recognized. For provenance, the review reports that it conducted a reverse-image search and found the image had begun circulating recently. It then traced an earlier appearance to the official Reddit group associated with MidJourney, an AI image-generation system. On that basis, the review concluded that the image was not an old photograph and was very likely AI-generated. It also reports finding no credible records that such a giant person lived in 1901; the supplied excerpt does not identify the archives or databases searched for that point. The review additionally describes visual concerns raised by AI researchers quoted by Reuters. Andrew Chen of the University of Auckland pointed to unclear, distorted faces. Ricardo Yates of Northeastern University cited irregular background colors and image resolution, indistinct eyes, and shadows that did not appear consistent or natural. The review’s conclusion was that the circulated image was very likely AI-generated, not a photograph of a real giant. The article cautions that visual inspection and detection software are not conclusive on their own. It recommends checking an image’s source and looking closely for distortions, unnatural anatomy, smoothness, or odd background details, while noting that such clues can suggest AI generation but cannot establish it with certainty.

El Detector / Univision NoticiasTechnology & AI

Video muestra conversación con un androide o robot creado con inteligencia artificial.

Publisher’s rating: Falso

The claim concerns a viral Facebook video that appears to show Chloe, presented as a domestic assistant, speaking with a male voice-over. The review says the post’s accompanying comment suggested that people would become robots created with artificial intelligence, and explains that the footage does not show a real android or AI-created robot: Chloe is a character in the video game Detroit: Become Human. The game was launched in May 2018 by Quantic Dream. (The review describes the clip and the claim as follows: “un video que aparentemente muestra una conversación con una androide llamada Chloe”; it concludes that “Chloe, como se identifica en la grabación, es un personaje del videojuego Detroit: Become Human.”) To investigate the clip, the review reports using an English-language Google keyword search, “Chloe android robot,” and finding that Chloe is a game character. It also compared the viral footage with an official promotional video published in 2018 on PlayStation Europe’s YouTube account. The review says the images and conversation match that official video, establishing the footage’s source rather than merely showing a similar robot or conversation. (It reports that “las imágenes objeto de esta verificación son las mismas que se muestran en el video oficial, así como también la conversación entre Chloe y la voz en off.”) The review identifies the missing context that made the clip appear to depict a real android: the Facebook post omitted the opening seconds that reveal it is a video game and the ending that announces the game’s release date. It therefore concludes that the viral video is promotional footage for the game, not evidence of a real robot created with AI.

dpaTechnology & AI

Klaus Schwab huurt miljoenen informatiestrijders in om de controle over het internet te grijpen.

Publisher’s rating: De bewering is gebaseerd op een vals artikel van Frontnieuws over een podcast van het WEF. In die podcast vertelt een woordvoerder van de VN over een programma van de VN om misinformatie tegen te gaan. Er hadden zich destijds 110.000 vrijwilligers aangemeld voor dat programma. Het WEF heeft hier niets mee te maken. Een tweet van het WEF over het zuiveren van het internet is nep.

LupaTechnology & AI

Nota Oficial O Google, Meta, Twitter, Telegram e Spotify uniram-se em uma iniciativa inédita para defender a liberdade de expressão e a democracia no Brasil [...] De maneira conjunta e com pesar, informamos que no dia 4 de Julho de 2023, à 00h01, suspenderemos as atividades no Brasil de todos os produtos e serviços ofercidos por esse grupo como forma de trazer atenção ao grande público, ao que consideramos, um rápido processo de destruição da democracia brasileira, a segunda maior do mundo ocidental.

Publisher’s rating: Falso