OpenAI Secures Key Legal Victory as Copyright Lawsuit Dismissed, Google Faces Judicial Criticism Over Evidence Handling in Monopoly Cases, U.S. Orders TSMC to Halt AI Chip Shipments to China
Welcome to Hashtag Trending. I’m your host, Jim Love. Let’s get into it.
Google Faces Judicial Criticism Over Evidence Handling in Monopoly Cases
Google is under fire in three major anti-monopoly lawsuits for allegedly destroying internal communications to avoid providing evidence. Judges in federal courts, including in Virginia, California, and the District of Columbia, have sharply criticized the tech giant’s handling of internal chats and communications, accusing it of suppressing evidence critical to the cases.
In a high-profile incident, a Google director, while discussing potential legal risks, expressed concern about a chat history being recorded, which could later be accessed by legal opponents. This example, along with others, was submitted as evidence of Google’s attempts to conceal information. Judges in these cases have noted Google’s practice of setting chats to auto-delete and misusing attorney-client privilege to shield internal discussions from legal scrutiny.
In California, Judge James Donato described Google’s actions as “disturbing,” accusing the company of creating a culture of concealment. He indicated that sanctions would follow due to Google’s deletion of chats in the case involving its Play Store monopoly. In Virginia, Judge Leonie Brinkema echoed similar concerns, accusing Google of using lawyers as a “smokescreen” to prevent evidence from being revealed. Meanwhile, in the District of Columbia, Judge Amit Mehta criticized Google’s practices but refrained from issuing sanctions, though he warned that the company may not be as fortunate in future cases.
Google maintains that it takes evidence preservation seriously and argues that it has made efforts to comply with legal requirements, but the courts remain skeptical of its practices.
U.S. Orders TSMC to Halt AI Chip Shipments to China
The U.S. government has ordered Taiwan Semiconductor Manufacturing Co. (TSMC) to stop shipping advanced chips to Chinese customers, effective Monday. According to a source, the Department of Commerce imposed export restrictions on chips with 7-nanometer or more advanced designs, which are commonly used in artificial intelligence (AI) accelerators and graphics processing units (GPUs). This move is part of broader efforts to prevent AI technology from reaching Chinese companies like Huawei, which has been a focal point of U.S. trade restrictions.
The order follows a recent report that one of TSMC’s chips was discovered in Huawei’s AI processor, in what appeared to be a violation of U.S. export controls. The U.S. government has expressed concern that such chips could enhance China’s AI capabilities, leading to tightened export rules.
TSMC has suspended shipments to affected Chinese clients, including Sophgo, after one of its chips was found in Huawei’s Ascend 910B AI processor. The discovery has raised alarms about the potential diversion of restricted technologies to Huawei and other Chinese firms.
The U.S. Commerce Department’s directive was issued through an “is informed” letter, which allows for swift action without formal rule-making procedures. This latest move intensifies existing trade tensions, as U.S. lawmakers from both parties push for stronger enforcement of export controls on China.
OpenAI Secures Key Legal Victory as Copyright Lawsuit Dismissed
A New York federal judge has dismissed a copyright lawsuit filed against OpenAI by news outlets Raw Story and AlterNet, delivering a significant legal win for the AI company. The case centered on OpenAI’s use of news articles as training data for its language models, but the court found that the plaintiffs failed to demonstrate concrete harm.
Judge Colleen McMahon ruled in favor of OpenAI’s fair use defense, highlighting that ChatGPT’s responses are synthesized from the training data rather than being direct copies of the articles. She noted that the risk of exact replication of articles is minimal, and factual information in the news isn’t subject to copyright protection.
The plaintiffs had focused on the removal of copyright management information rather than direct copyright infringement, but McMahon found that this did not show sufficient harm.
While she dismissed the case, she left open the possibility for an amended complaint, though expressed doubt about the plaintiffs’ ability to prove significant injury.
This ruling could be a big deal and have big implications for similar lawsuits, including ongoing cases involving AI-generated content and copyright claims, such as the New York Times’ case against OpenAI. By affirming that AI-generated content is a synthesis rather than a reproduction of copyrighted material, the court has strengthened OpenAI’s defense in the broader legal landscape surrounding AI training practices.
For those who produce news content, it’s a blow. It’s already hard to get money to pay for journalism. But equally, for those who are interested in the advancement of generative AI, it’s hard to imagine that we would be as far as we are if the existing models hadn’t been able to have that ready made training set. And of course, any smaller company or open-source model would never be able to pay for the data they would need.
I don’t have any answers. Sure like to hear your opinions on this one.
And that’s our show for today.
Reach me at editorial@technewsday.ca
I’m your host Jim Love, have a Terrific Tuesday
