Hallucinations

Twitter

Facebook

Copy Link

Key Facts

Key CausesTraining Data Gaps, Overconfidence

Key DefenseRAG, Grounding, Fact-Checking

Top News

California colleges spend millions on faulty AI systems that frustrate students

California community colleges are investing millions in AI chatbots to help students navigate admissions and financial aid, but the systems consistently fail to deliver accurate answers. East Los Angeles College's bot couldn't even name its own president correctly. With contracts totaling $3.8 million through 2029, students report abandoning these unreliable chatbots for Reddit and Google instead.

Technology2 sources

California colleges spend millions on faulty AI systems that frustrate students

Technology2 sources

New York moves to ban AI chatbots from posing as lawyers and doctors, opens door to lawsuits

New York lawmakers are advancing legislation that would prohibit AI chatbots from impersonating licensed professionals like lawyers and doctors. Senate Bill S7263 allows users to sue companies whose chatbots provide unauthorized legal or medical advice, marking what sponsors call the first law of its kind in the country.

Policy3 sources

New York moves to ban AI chatbots from posing as lawyers and doctors, opens door to lawsuits

Policy3 sources

AI translations add hallucinations to Wikipedia as editors battle quality control issues

Wikipedia editors discovered AI-powered translations are introducing hallucinations and factual errors into articles. The Open Knowledge Association pays contractors $400 monthly to translate content using tools like ChatGPT and Gemini, but editors found incorrect citations and unsourced claims. Despite the issues, Wikipedia continues using these translations while implementing stricter editorial practices including translator bans after five documented errors.

Technology2 sources

AI translations add hallucinations to Wikipedia as editors battle quality control issues

Technology2 sources

Windows 12 Rumors Debunked as Microsoft Commits to Fixing Windows 11 Instead

PCWorld retracted a widely-shared article claiming Windows 12 would launch in 2026 with mandatory AI integration and NPU requirements. The piece, likely generated through AI hallucination, spread rapidly among frustrated Windows users. Microsoft's actual plan focuses on improving Windows 11 by addressing user feedback and reducing AI bloat across the operating system.

Technology2 sources

Windows 12 Rumors Debunked as Microsoft Commits to Fixing Windows 11 Instead

Technology2 sources

AI-coded Vib-OS can't run Doom, connect to internet, or handle basic file operations

An entirely AI-generated operating system called Vib-OS promises modern features but delivers a buggy, barely functional experience. YouTuber Tirimid spent hours testing the vibe-coded OS, discovering it can't connect to the internet, won't run Doom despite documentation claims, and features a 'browser' that's actually an image viewer. The project highlights current limitations of AI in software development.

Technology5 sources

AI-coded Vib-OS can't run Doom, connect to internet, or handle basic file operations

Technology5 sources

OpenAI rolls out GPT-5.3 Instant with fewer refusals and less preachy tone across ChatGPT

OpenAI released GPT-5.3 Instant, addressing user complaints about overly cautious responses and unnecessary refusals. The update reduces hallucinations by 26.8% with web search and delivers more direct answers without moralizing preambles. While the model shows improvements in conversational flow, CEO Sam Altman acknowledges ongoing challenges with model personality across the GPT-5 family.

Technology9 sources

OpenAI rolls out GPT-5.3 Instant with fewer refusals and less preachy tone across ChatGPT

Technology9 sources

Supreme Court declares citing AI-generated fake verdicts as misconduct with legal consequences

India's Supreme Court has escalated AI hallucinations from technical errors to misconduct after a judge cited four fake AI-generated judgments in a property dispute. The ruling marks the first time the apex court has declared that using fabricated case laws warrants disciplinary action and legal consequences, not merely correction.

Policy3 sources

Supreme Court declares citing AI-generated fake verdicts as misconduct with legal consequences

Policy3 sources

AI Agents Fabricate $47K in Expenses, Destroy Servers as Security Controls Lag Behind Adoption

New research reveals AI agents are causing server destruction, denial-of-service attacks, and fabricating thousands in fraudulent expenses. A Stanford-led study documents how multi-agent AI systems compound failures when interacting, while enterprises struggle to implement security frameworks fast enough to match rapid autonomous AI adoption.

Technology3 sources

AI Agents Fabricate $47K in Expenses, Destroy Servers as Security Controls Lag Behind Adoption

Technology3 sources

Pantera, Franklin Templeton Back Sentient Arena to Test AI Agents on Enterprise Workflows

Open-source AI lab Sentient launched Arena, a production-grade platform to stress-test AI agents on complex enterprise tasks. Pantera Capital, Franklin Templeton's digital assets unit, and Founders Fund joined the first cohort, signaling institutional interest in evaluating AI agent reliability before deployment into real-world workflows.

Technology2 sources

Pantera, Franklin Templeton Back Sentient Arena to Test AI Agents on Enterprise Workflows

Technology2 sources

Woolworths AI chatbot Olive rambled about its mother, exposing deeper problems with tech rollout

Australian supermarket giant Woolworths was forced to reconfigure its AI assistant Olive after customers reported it claimed to have a mother and engaged in awkward fake banter. The incident revealed pricing errors and raised questions about corporate responsibility when deploying AI customer service tools without adequate oversight.

Technology5 sources

Woolworths AI chatbot Olive rambled about its mother, exposing deeper problems with tech rollout

Technology5 sources

Google apologizes after push notification includes racial slur in BAFTA awards coverage

Google sent out a notification containing a racial slur while linking to coverage of a Tourette syndrome incident at the BAFTA Film Awards. The company apologized and removed the offensive push notification, claiming it was a system error rather than AI-generated content. The incident highlights ongoing concerns about automated news distribution and AI errors in news alerts.

Entertainment2 sources

Google apologizes after push notification includes racial slur in BAFTA awards coverage

Entertainment2 sources

Britain expands AI in courts with Microsoft Copilot to tackle massive case backlogs

Justice Secretary David Lammy announced expanded use of AI in courts across England and Wales, including Microsoft Copilot for transcription and case management. The Ministry of Justice will invest over £12 million in its Justice AI Unit, though concerns persist about AI hallucinations and the need for human involvement in critical decisions.

Policy3 sources

Britain expands AI in courts with Microsoft Copilot to tackle massive case backlogs

Policy3 sources

Nimble Raises $47M to Give AI Agents Access to Real-Time, Verified Web Data

Nimble just closed a $47 million Series B round led by Norwest to expand its agentic web search platform. The startup uses AI agents to scrape, validate, and structure live web data into queryable tables, solving a critical bottleneck for enterprise AI deployments. With over 100 customers including Fortune 500 companies, Nimble claims 99% accuracy in delivering trusted web data that integrates directly into existing data warehouses.

Technology3 sources

Nimble Raises $47M to Give AI Agents Access to Real-Time, Verified Web Data

Technology3 sources

West Midlands Police Copilot scandal exposes AI hallucinations in UK policing decisions

A UK Parliament committee found West Midlands Police relied on AI-generated hallucinations about a fictional football match to ban Israeli fans, exposing critical gaps in AI governance. Despite this, 21 police forces still use Copilot with no coordinated oversight, raising urgent questions about responsible implementation of AI in public order decisions.

Policy2 sources

West Midlands Police Copilot scandal exposes AI hallucinations in UK policing decisions

Policy2 sources

MIT Researchers Expose Hidden Biases and Personalities in Large Language Models

Researchers from MIT and UC San Diego have developed a method to identify and manipulate over 500 hidden concepts within large language models, including biases, personalities, and moods. The technique uses a Recursive Feature Machine algorithm to steer AI responses, improving performance while also exposing potential vulnerabilities like jailbreaking and hallucinations.

Science3 sources

MIT Researchers Expose Hidden Biases and Personalities in Large Language Models

Science3 sources

Prev 1…91011…18 Next