OpenAI has evidence that its models helped train China’s DeepSeek

Sucking in data you didn’t ask permission for? Sounds familiar.

Chinese artificial intelligence company DeepSeek disrupted Silicon Valley with the release of cheaply developed AI models that compete with flagship offerings from OpenAI — but the ChatGPT maker suspects they were built upon OpenAI data.

OpenAI and Microsoft are investigating whether the Chinese rival used OpenAI’s API to integrate OpenAI’s AI models into DeepSeek’s own models, according to Bloomberg. The outlet’s sources said Microsoft security researchers detected that large amounts of data were being exfiltrated through OpenAI developer accounts in late 2024, which the company believes are affiliated with DeepSeek.

OpenAI told the Financial Times that it found evidence linking DeepSeek to the use of distillation — a common technique developers use to train AI models by extracting data from larger, more capable ones. It’s an efficient way to train smaller models at a fraction of the more than $100 million that OpenAI spent to train GPT-4. While developers can use OpenAI’s API to integrate its AI with their own applications, distilling the outputs to build rival models is a violation of OpenAI’s terms of service. OpenAI has not provided details of the evidence it found.

The situation is rich with irony. After all, it was OpenAI that made huge leaps with its GPT model by sucking down the entirety of the written web without consent.

President Donald Trump’s artificial intelligence czar David Sacks said “it is possible” that IP theft had occurred. “There’s substantial evidence that what DeepSeek did here is they distilled knowledge out of OpenAI models and I don’t think OpenAI is very happy about this,” Sacks told Fox News on Tuesday.

“We know PRC (China) based companies — and others — are constantly trying to distill the models of leading US AI companies,” OpenAI said in a statement to Bloomberg. “As the leading builder of AI, we engage in countermeasures to protect our IP, including a careful process for which frontier capabilities to include in released models, and believe as we go forward that it is critically important that we are working closely with the US government to best protect the most capable models from efforts by adversaries and competitors to take US technology.”

from The Verge https://ift.tt/A83ZhQI

Comments

Pandora Stories lets artists add commentary to their own playlists

Pandora launched Stories today, a tool that lets artists and creators add voice commentary to their own playlists. The Stories feature merges podcasts with music playlists, and is meant for artists to add context to an album, or for podcasters to experiment with new storytelling formats. The feature is part of Pandora AMP, the streaming service’s free Artist Marketing Platform that helps creators promote their work. To kick off the launch, Pandora’s prepared some Stories by artists like John Legend and Daddy Yankee, who tell listeners their personal stories interspersed between their own songs. There’s also a Stories playlist called Love Songs That Aren’t Really Love Songs , which includes commentary on individual songs like a podcast... Continue reading… from The Verge - All Posts https://ift.tt/2Xz1oNc

Minneapolis hiring social media influencers for former police officers’ trials

George Floyd died in police custody in May 2020 | Getty Minneapolis is hiring social media influencers to share “city-generated and approved messages” during the trials of four former city police officers charged in the May 2020 killing of George Floyd, according to the Minnesota Reformer . The Minneapolis city council approved the plan on Friday, which calls for six influencers to be paid about $2,000 each to spread the city’s messages with Black, Native American, Hmong, and Latino communities. The goal is to “address/dispel incorrect information” by using “trusted messengers,” part of a program the city refers to as its Joint Information System to share “timely and relevant information” with the public during the trials. “Through the Communications and Neighborhood and Community Relations... Continue reading… from The Verge - All Posts https://ift.tt/3q1AY3x

Android Addition Opens FIDO Password Killer to Billions

The FIDO Alliance hammered another nail into the passwords coffin with the announcement that devices running Android 7.0 or higher will be compatible with FIDO2. Certification of Android 7.0+ means devices running those versions of Google's mobile operating system will support FIDO2 out of the box or through a software update. FIDO2, introduced last year, provides a FIDO Web authentication standard that combines the World Wide Web Consortium's Web Authentication specification with FIDO's Client-to-Authenticator protocol. from TechNewsWorld https://ift.tt/2GQVG4f

Latest Technology

Search This Blog