Data Activists Target OpenAI In Challenge To ChatGPT’s ‘Hallucination’ Problem

Privacy activists filed a complaint against ChatGPT maker OpenAI on Monday over the company’s failure to correct misinformation its chatbot regularly “hallucinates” about people, a move that will increase pressure on tech firms to address what has become a well known but difficult-to-fix problem as they race to roll out the AI tools to more customers.

“Simply making up data about individuals is not an option,” the group said in a statement about the complaint, which was filed on behalf of an unnamed “public figure” and accuses OpenAI of refusing to correct or erase false information and statements it had made up about the individual. For example, the group said ChatGPT gave “various inaccurate information” when asked about the figure’s birth date and that OpenAI said “there is no way to prevent its systems” from displaying the false information.

Instead, the complaint said OpenAI only offered to block or filter results based on prompts like the figure’s name — which would filter all information about them — something noyb said would still leave the incorrect data in OpenAI’s systems, “just not shown to users.” It also accused OpenAI of failing to disclose relevant information about the person when requested, including what data had been processed, its sources and who it had been shared with, a legal obligation noyb lawyer Maartje de Graaf said “applies to all companies,” adding that it is “clearly possible” to keep track of information sources when training an AI system.“Making up false information is quite problematic in itself.

Source: Healthcare Press (healthcarepress.net)

Chatgpt Noyb Schrems Gdpr Hallucinate Hallucination Misinformation Data Privacy Complaint

United States Latest News, United States Headlines

Similar News:You can also read news stories similar to this one that we have collected from other news sources.

Data Activists Target OpenAI In Challenge To ChatGPT’s ‘Hallucination’ ProblemI am a senior reporter for the Forbes breaking news team, covering health and science from the London office. Previously I worked as a reporter for a trade publication covering big data and law and as a freelance journalist and policy analyst covering science, tech and health.
Source: ForbesTech - 🏆 318. / 59 Read more »

OpenAI’s rules can be ‘easily’ dodged to target Latinos, study warnsChatGPT can be used to ‘maliciously’ court minorities, researchers say.
Source: washingtonpost - 🏆 95. / 72 Read more »

Managing Large Data Volumes With MinIO, Langchain and OpenAIA practical guide to integrating MinIO, Langchain and OpenAI’s GPT-3.5 model focusing on summarizing documents stored in MinIO buckets.
Source: hackernoon - 🏆 532. / 51 Read more »

Microsoft And OpenAI Partnering On Stargate, A $100B U.S. Data CenterDr. Cindy Gordon ICD.D. is a CEO, a thought leader, author, keynote speaker, board director, and advisor to companies and governments striving to modernize their business operations, with advanced AI methods. She is the CEO and Founder of SalesChoice, an AI SaaS company focused on ending revenue uncertainty and human advantage.
Source: ForbesTech - 🏆 318. / 59 Read more »

Climate activists have new target in mind for major summer protest: 'Make their lives miserable'A group of climate activists is calling for a “historic blockade' and to 'act now, boldly and swiftly' during an event over the summer.
Source: FoxNews - 🏆 9. / 87 Read more »

OpenAI transcribed over a million hours of YouTube videos to train GPT-4A new report details how OpenAI, Google, and Meta have tried to get over the limits of good data available online — for OpenAI, that includes generating YouTube transcripts.
Source: verge - 🏆 94. / 67 Read more »