Mealey's Artificial Intelligence

  • January 12, 2024

    Stanford Researchers Find ‘Alarmingly High’ Failures In Legal AI

    STANFORD, Calif. — Large language models (LLMs) demonstrated an “alarmingly high” error rate when used in the legal field, with the AIs unable to identify core holdings in opinions at least 75% of the time and producing hallucinations at rates of up to 88% when they were asked specific legal inquiries, Stanford University researchers said Jan. 11.

  • January 11, 2024

    AI Can’t Patent Inventions, United Kingdom’s Supreme Court Says

    LONDON — The law of the United Kingdom recognizes natural persons as inventors and does not extend to an artificial intelligence, the country’s top court said in affirming denial of two patents.

  • January 10, 2024

    Clothing Retailer Rejects Claims Its AI Systematically Steals

    LOS ANGELES — Claims involving a clothing retailer’s use of artificial intelligence implicate ordinary business practices and cannot form the basis of Racketeer Influenced and Corrupt Organizations Act and copyright claims, Shein Distribution Corp. says in a reply brief rebutting designers’ claims that the company systematically duplicates and steals protected and commercially valuable designs.

  • January 10, 2024

    Plaintiffs Amend Claims Accusing Google Of Data ‘Theft’ To Train AI Chatbot

    SAN FRANCISCO — A group of previously anonymous putative class action plaintiffs filed an amended complaint in California federal court identifying themselves as a New York Times bestselling author and several users of Google LLC services, all now accusing Google of “data theft” and violation of California’s unfair competition law (UCL), copyright laws and other statutes by using their data to train its artificial intelligence chatbot known as “Bard.”

  • January 10, 2024

    Pennsylvania, OpenAI Partnership Puts AI In Administrative Office

    HARRISBURG, Pa. — Pennsylvania will employ an enterprise version of ChatGPT for use in its Office of Administration, Gov. Josh Shapiro said Jan. 9 in what a press release called a “first-in-the-nation” pilot program.

  • January 10, 2024

    FTC Warns Of Potential Privacy Pitfalls For AI Companies

    WASHINGTON, D.C. — The Federal Trade Commission in a Jan. 9 announcement reminded artificial intelligence companies that they could face regulatory actions if their large language models’ (LLM) voracious demand for data tramples obligations to protect users’ data and privacy, either through the improper use of data or how the company handles representations about what it does with the data.

  • January 09, 2024

    Microsoft, OpenAI Face Journalists’ Suit Over AI Training

    NEW YORK — Microsoft Corp. and various OpenAI are no better than any run-of-the-mill thief after they “systematically pilfered” copyrighted works despite enjoying both the means and ability to pay, two journalists allege in a class action filed in federal court in New York.

  • January 04, 2024

    In AI Information Request Case, Justice Upholds ‘Snail Mail’ Appeal Rule

    NEW YORK — Even in an era where parties communicate through email about appeals in a case requesting information about an transportation agency’s utilization of artificial intelligence, the government can still require that the actual appeal be filed through traditional mail services, a New York justice said in denying relief in a Freedom of Information Law (FOIL) case.

  • January 02, 2024

    AI Generated Fake Cites, Trump-Adjacent Attorney Michael Cohen’s Lawyer Says

    NEW YORK — While attempting to secure an early exit from supervised release, Michael Cohen used fake cites he found through Google Inc.’s Bard without realizing that the program was not a “super-charged” search engine but rather a generative artificial intelligence prone to error, according to a letter response filed by Cohen’s lawyer in the U.S. District Court for the Southern District of New York.

  • January 02, 2024

    Latest AI Copyright Suit Sees New York Times Target Microsoft, OpenAI

    NEW YORK — The New York Times Co. (NYT) sued ChatGPT owners Microsoft Corp. and various OpenAI entities for copyright infringement, saying there is nothing transformative about using valuable and protected materials to create a substitute product.

  • January 02, 2024

    Committee Gets OK For Vote On Liquidation Plan In Vesttoo Chapter 11 Cases

    WILMINGTON, Del. — A Delaware federal bankruptcy judge who is allowing the Official Committee of Unsecured Creditors to put its Chapter 11 plan of liquidation for Vesttoo Ltd. and 48 affiliates to a vote also granted its motions for leave to conduct discovery against two banking organizations.

  • December 22, 2023

    Former Governor, Others Sever, Transfer, Dismiss Portions Of AI Copyright Suit

    NEW YORK — Former Arkansas Gov. Mike Huckabee and other copyright holders who filed suit over the use of their works to train artificial intelligence told a federal judge in New York that they reached an agreement to dismiss one defendant and sever and transfer their claims against Microsoft Corp. and Meta Platforms Inc. to a federal court in California and said they would not oppose a fourth defendant’s motion to stay discovery while the court resolves pending motions.

  • December 20, 2023

    FTC Places 5-Year Ban On Rite Aid Using AI Facial Recognition Software

    WASHINGTON, D.C. — Rite Aid Corp.’s artificial intelligence facial recognition software generated thousands of false matches, often targeting women or people of color, the Federal Trade Commission said Dec. 19 in announcing that it was imposing a five-year ban on the retailer’s use of the technology.

  • December 19, 2023

    Study: AI Outperforms Human Clinicians In Limited Settings

    BOSTON — Large language model (LLM) artificial intelligence (AI) ChatGPT-4 performed better than humans at diagnosing five diseases in pre- and post-test situations, but only when the test results were negative, according to a new peer-reviewed study released on the JAMA Network.

  • December 19, 2023

    Unsecured Creditors Update Proposed Liquidation Plan In Vesttoo Chapter 11 Cases

    WILMINGTON, Del. — The Official Committee of Unsecured Creditors on Dec. 18 filed an amended proposed combined disclosure statement and Chapter 11 plan of liquidation for the jointly administered cases of Vesttoo Ltd. and 48 affiliates in Delaware federal bankruptcy court.

  • December 15, 2023

    RIAA To High Court: Use Equitable Tolling For ‘Black Box,’ AI-Based Infringement

    WASHINGTON, D.C. — In one of eight amicus curiae briefs filed in a dispute over the discovery accrual rule in the context of copyright infringement, the Recording Industry Association of America (RIAA) urges the U.S. Supreme Court to clarify that equitable tolling applies to the Copyright Act’s three-year limitations period when the infringement “is undetectable as a practical matter.”

  • December 15, 2023

    Plaintiffs:  Humana Uses Artificial Intelligence, Denies Senior Care In Bad Faith

    LOUISVILLE, Ky. — Humana Inc. denies elderly patients Medicare Advantage Plan coverage to which they are entitled through the employment of artificial intelligence in place of actual doctors, plaintiffs claim in a class action filed in a Kentucky federal court.

  • December 14, 2023

    Journalist Says Government Can’t Defend Withholding FOIA AI Documents

    WASHINGTON, D.C. — No statutory or logical basis prevents the Department of Defense (DOD) from releasing 5,000 documents it concedes are responsive to a Freedom of Information Act (FOIA) inquiry about Google Inc.’s role in the Project Maven artificial intelligence program, a journalist tells a federal judge in the District of Columbia in a cross-motion and opposition to summary judgment.

  • December 14, 2023

    Israeli Law Raised In Feedback On Liquidation Plan For Vesttoo And Its Affiliates

    WILMINGTON, Del. — In two Dec. 13 filings citing Israeli law and other issues, Vesttoo Ltd. and its affiliated debtors and a venture capital organization and individual that report significant related interests urge a Delaware federal bankruptcy court not to approve a Chapter 11 plan of liquidation and related arrangements as currently proposed.

  • December 13, 2023

    Authors Amend Copyright Claim Against Meta For Using Books To Train AI

    SAN FRANCISCO — Three weeks after a California federal judge called their theory of copyright infringement “nonsensical,” 13 authors filed an amended putative class complaint against Meta Platforms Inc. dropping all claims except for a direct copyright infringement claim against Meta for using their copyrighted works to train its AI chatbots.

  • December 12, 2023

    State Court Hears Arguments On Motion To Dismiss In ChatGPT Defamation Case

    ATLANTA — A Georgia state court held a hearing on a OpenAI LLC’s motion to dismiss defamation claims against it over ChatGPT content naming a journalist as a defendant in a financial misconduct case to which he was not a party.

  • December 12, 2023

    SEC Says AI Startup Is Wrong To Call False Statements About Legality Unactionable

    ORLANDO, Fla. — An artificial intelligence startup and its founder are incorrect to say the Securities and Exchange Commission’s complaint against them are based on unactionable statements, the SEC says in a Dec. 11 memorandum filed in a federal court in Florida in opposition to the motion to dismiss, saying the complaint sufficiently alleges that the company and its founder lured investors with false promises of early access to profits from an AI “ecosystem.”

  • December 12, 2023

    Parties To AI Rental Price Antitrust Suit Brief Recent Precedent

    NASHVILLE, Tenn. — With a motion to dismiss fully briefed and the United States filing a notice of interest in a case alleging antitrust violations stemming from the use of algorithmic and artificial intelligence-based rental pricing software, the parties briefed a federal judge in Tennessee on the impact of recent Fourth Circuit U.S. Court of Appeals precedent governing the requirements for antitrust claims.

  • December 12, 2023

    Washington High Court Seeks Comments On Attorney Training, Including AI Chatbot Use

    SEATTLE — The Washington Supreme Court asked for comments about proposed amendments to the state’s continuing legal education system involving legal technology, including the handling and use of artificial intelligence chatbots.

  • December 12, 2023

    Judge Stays Discovery In Case Alleging AI Founder Was Tricked Into Selling Stake

    SAN FRANCISCO — A federal judge in California stayed discovery pending resolution of a motion to dismiss in a man’s case claiming that he was tricked into believing that his stake in an artificial intelligence company was worthless.