Discord fixes AI moderation bug that wrongfully banned users for harmless images

·37 views

Discord has acknowledged a bug in its AI-powered moderation system that resulted in users being wrongfully banned. The issue had been silently affecting accounts since May, going undetected for several months. Over a single weekend, an additional 200 users were incorrectly banned before Discord's team caught the problem. The company has since identified the root cause and deployed a fix to prevent further wrongful bans.

Read the full story at TechCrunch

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

UK to Invest £100m in AI Biosecurity Lab Ahead of Budget

The UK government plans to spend £100 million on an artificial intelligence biosecurity laboratory, which would be described as a world first. The chancellor is expected to announce the funding in Wednesday's Budget. The investment is intended to strengthen the country's defences against biological threats. The source article provides limited detail beyond this, so further specifics on the lab's structure and timeline are not available.

0 comments Read more at BBC Tech

TechnologyTechCrunch ·

Startups Can Win Customers Without Going Viral, Panelists Say

At TechCrunch Disrupt last week, two founders and a venture capitalist held a discussion on building a startup's customer base. The panel focused on how companies can attract an early following. Speakers also addressed how startups can keep those early users engaged over time. The session was framed around the idea that viral reach is not required for success.

0 comments Read more at TechCrunch

TechnologyTechCrunch ·

OpenAI Says It Will Stop Training Its AI Models on the Live Internet

OpenAI has announced it will no longer train its frontier AI models on live internet data, citing significant risks. The decision reverses an earlier position, arriving only weeks after one of the company's engineers had argued that such a restriction would not be workable. The source article provides limited detail beyond this shift in policy. The announcement highlights ongoing debate within the AI industry over how training data should be sourced and safeguarded.

0 comments Read more at TechCrunch

TechnologyArs Technica ·

NASA Revises Lunar Lander Plans and Resolves a Longstanding Orion Issue

NASA has announced updated plans for its Moon lander program, according to a report from Ars Technica. The agency says it has the mandate, resources, and stakeholder backing needed to move forward. The revised approach is also said to address a long-running problem involving the Orion spacecraft. The article provided is only a single quoted line, so further specifics about timelines, costs, or technical details are not available from this source. Readers should consult the full original reporting for complete information.

0 comments Read more at Ars Technica