Última hora
RUФСБ предотвратила теракт на военном аэродроме в РостовеRUUAV Attack Prompts Evacuations in Taganrog, RussiaRUФСБ предотвратила атаку Киева на аэродром "Ростов-Центральный"KR더불어민주당 임오경 의원, 축구협회 청문회 참고인 신청 철회DEWirtschaftsverbände wehren sich gegen Abschaffung von MinijobsRUФСБ: Киев планировал атаку на аэродром "Ростов-Центральный" дронамиKR파키스탄, 탈레반 갈등으로 이란·중국 통과 우회 무역로 2개 개설RUФСБ предотвратила крупный теракт на аэродроме в Ростовской областиDEDAX stagniert trotz Tech-Rally – US-Angriffe auf Iran verunsichern Anleger kaumCN華航6月、Q2、H1營收同創歷史新高RUФСБ предотвратила теракт на военном аэродроме в РостовеRUUAV Attack Prompts Evacuations in Taganrog, RussiaRUФСБ предотвратила атаку Киева на аэродром "Ростов-Центральный"KR더불어민주당 임오경 의원, 축구협회 청문회 참고인 신청 철회DEWirtschaftsverbände wehren sich gegen Abschaffung von MinijobsRUФСБ: Киев планировал атаку на аэродром "Ростов-Центральный" дронамиKR파키스탄, 탈레반 갈등으로 이란·중국 통과 우회 무역로 2개 개설RUФСБ предотвратила крупный теракт на аэродроме в Ростовской областиDEDAX stagniert trotz Tech-Rally – US-Angriffe auf Iran verunsichern Anleger kaumCN華航6月、Q2、H1營收同創歷史新高
Newsgather
BackDiscord Bug Led to Thousands of Accounts Being Incorrectly Banned
Discord Bug Led to Thousands of Accounts Being Incorrectly Banned
En desarrollo
Engadget2 g önceTecnología

Discord Bug Led to Thousands of Accounts Being Incorrectly Banned

En resumen

  • A bug in Discord's moderation tools mistakenly banned around 8,200 accounts since May 2026, with about 200 affected this past weekend.
  • The issue, caused by AI flagging square grid images as child sexual abuse material, led to erroneous bans that were not automatically lifted.

Resumen generado por IA

Tamaño de fuente

A bug in Discord's moderation tools mistakenly banned around 8,200 accounts since May 2026, with about 200 affected this past weekend. The issue, caused by AI flagging square grid images as child sexual abuse material, led to erroneous bans that were not automatically lifted.

Read the full article on Engadget

Temas relacionados

This article was originally published by Engadget.

Noticias relacionadas

Telstra Admits Knowing Time System Dangers, Investigates Backup Failures After Outage
En desarrollo·2 sa önce

Telstra Admits Knowing Time System Dangers, Investigates Backup Failures After Outage

Telstra is investigating why its backup systems failed during a recent national outage caused by a software fault in its time systems. The outage impacted mobile, Eftpos, and rail services for nearly five hours. South Australian police are investigating a death during the outage, exploring a potential link to triple zero call failures, though Telstra claims no record of such a failure.

Guardian World
OpenAI Engineer Leads ChatGPT Transformation, Cursor's AI Model Future, Meta's AI Chaos, and More Tech News
En desarrollo·6 sa önce

OpenAI Engineer Leads ChatGPT Transformation, Cursor's AI Model Future, Meta's AI Chaos, and More Tech News

OpenAI's Thibault Sottiaux is leading a major ChatGPT overhaul, while Cursor navigates its future with third-party AI models post-SpaceX acquisition. Meta faces internal turmoil over its AI strategy and employee tracking, and Google offers Workspace promo codes. Anthropic emphasizes its role in AI safety, and Coherence Neuro tests cancer-detecting brain implants. OpenAI delays new models at White House request, and Qualcomm acquires chip startup Modular.

Wired
OpenAI Engineer Leads ChatGPT Transformation, Cursor's Future with AI Models, Meta's AI Chaos, and More Tech News
En desarrollo·6 sa önce

OpenAI Engineer Leads ChatGPT Transformation, Cursor's Future with AI Models, Meta's AI Chaos, and More Tech News

Thibault Sottiaux is leading a major ChatGPT overhaul. Cursor faces uncertainty with SpaceX acquisition regarding third-party AI models. Meta's AI strategy is reportedly chaotic, with employees disliking a company-wide hackathon. Google Workspace offers July 2026 promo codes. Anthropic believes its success is crucial for AI safety. Coherence Neuro is testing brain implants for cancer detection. Qualcomm acquired Modular for nearly $4 billion. OpenAI delayed GPT-5.6 models at the White House's request, and Meta paused an employee-tracking program after a data leak.

Wired
Más sobre este temadiscord