Hacker Exploits Anthropic Chatbot to Attack Mexican Gov't Agencies

A hacker exploited Anthropic PBC’s artificial intelligence chatbot to carry out a series of attacks against Mexican government agencies, resulting in the theft of a huge trove of sensitive tax and voter information, according to cybersecurity researchers. The unknown Claude user wrote Spanish-language prompts for the chatbot to act as an elite hacker, finding vulnerabilities in government networks, writing computer scripts to exploit them and determining ways to automate data theft, Israeli cybersecurity startup Gambit Security said in research.

Anthropic, Facing Competition, Scales Back Safety Commitment

Anthropic, the artificial-intelligence company known for its devotion to safety, is scaling back that commitment by softening its core safety policy to stay competitive with other AI labs. Anthropic previously paused development work on its model if it could be classified as dangerous, but said it would end that practice if a comparable or superior model was released by a competitor.

Apple Creating 'Age Assurance' Tools to Comply with Countries' Laws

Apple is launching new tools to comply with the growing number of age-verification laws both in the U.S. and abroad. As part of the changes, Apple will block the downloads of apps rated 18+ in Brazil, Australia, and Singapore, while also rolling out other features to comply with laws in Utah and Louisiana in the U.S. The company informed developers that it’s expanding its set of “age assurance” tools, including an updated Declared Age Range API now available for beta testing.

Defense Department Threatens to Cancel Anthropic's Contract

Defense Secretary Pete Hegseth gave Anthropic Chief Executive Dario Amodei three days to comply with the Pentagon’s demands on using its artificial-intelligence models or face cancellation of the company’s contract, people familiar with the matter said. If Anthropic doesn’t show more flexibility working with the military, Hegseth said he could also label the company a supply-chain risk, a move typically reserved for overseas companies linked to foreign adversaries, or invoke the Defense Production Act to essentially force the company to work more collaboratively with the Pentagon.

Hegseth Summons Anthropic CEO to Discuss Military's Use of Claude

Defense Secretary Pete Hegseth has summoned Anthropic CEO Dario Amodei to the Pentagon for what sources say is likely to be a tense meeting over terms for military use of Anthropic's Claude. Claude is the only AI model available in the military's classified systems, and the most capable model for sensitive defense and intelligence work. The Pentagon doesn't want to lose access to Claude but is furious with Anthropic for refusing to lift its safeguards entirely.

Anthropic Accuses Chinese Companies of Siphoning Info from Claude

U.S. artificial-intelligence startup Anthropic said three Chinese AI companies set up more than 24,000 fraudulent accounts with its Claude AI model to help their own systems catch up. The three companies — DeepSeek, Moonshot AI and MiniMax — prompted Claude more than 16 million times, siphoning information from Anthropic’s system to train and improve their own products, Anthropic said in a blog post.

Zuckerberg Defends Meta at Trial Over Social Media Addiction

In his first time testifying about child safety in front of a jury, Meta CEO Mark Zuckerberg said the company does not seek to make Instagram addictive to younger users, pushing back against claims that the social media app is designed to be harmful to children. “I’m focused on building a community that is sustainable,” he said when he was asked whether Meta wants people to be addicted to its social media platforms.

UK Prime Minister Wants to Fine Tech Firms for Not Removing 'Revenge Porn'

Deepfake nudes and “revenge porn” must be removed from the internet within 48 hours or technology firms risk being blocked in the UK, Keir Starmer has said, calling it a “national emergency” that the government must confront. Companies could be fined millions or even blocked altogether if they allow the images to spread or be reposted after victims give notice.

Disney Sends Demand Letter to ByteDance Over 'Seedance' AI Tool

The Walt Disney Company sent a cease-and-desist letter to ByteDance, alleging the Chinese tech giant has been infringing on its works to train and develop an AI video generation model without compensation, according to a copy of the letter obtained by Axios. It's the most serious action a Hollywood studio has taken so far against ByteDance since it launched Seedance 2.0.

Public Radio Host Sues Google for Recreating His Voice on NotebookLM

David Greene, a public radio veteran who has hosted NPR’s “Morning Edition” and KCRW’s political podcast “Left, Right & Center,” is suing Google, alleging that it violated his rights by building a product that replicated his voice without payment or permission, giving users the power to make it say things Greene would never say. The dispute involves NotebookLM, a language models trained on vast libraries of writing and speech by real humans who were never told their words and voices would be used in that way — raising profound questions of copyright and ownership.

Pentagon Considers Ending Work with Anthropic Over Pushback

The Pentagon is considering severing its relationship with Anthropic over the AI firm's insistence on maintaining some limitations on how the military uses its models, a senior administration official told Axios. The Pentagon is pushing four leading AI labs to let the military use their tools for "all lawful purposes," even in the most sensitive areas of weapons development, intelligence collection, and battlefield operations. Anthropic has not agreed to those terms, and the Pentagon is getting fed up after months of difficult negotiations.

DHS Send Subpoenas to Tech Companies for Info on ICE Protesters

The Department of Homeland Security is expanding its efforts to identify Americans who oppose Immigration and Customs Enforcement by sending tech companies legal requests for the names, email addresses, telephone numbers and other identifying data behind social media accounts that track or criticize the agency. In recent months, Google, Reddit, Discord and Meta, which owns Facebook and Instagram, have received hundreds of administrative subpoenas from the Department of Homeland Security, according to four government officials and tech employees privy to the requests.

WhatsApp Blocked for 'Unwillingness to Comply with Russian Law'

U.S. messenger app WhatsApp, owned by Meta Platforms, has been completely blocked in Russia for failing to comply with local law, the Kremlin said, suggesting Russians turn to a state-backed "national messenger" instead. "Due to Meta's unwillingness to comply with Russian law, such a decision was indeed taken and implemented," Kremlin spokesman Dmitry Peskov told reporters, proposing that Russians switch to MAX, Russia's state-owned messenger.

At Trial, Instagram CEO Says Social Media Not 'Clinically Addictive'

Instagram’s chief executive said that social media was not “clinically addictive,” disputing claims that the platform prioritized making money over the mental health of young users in a landmark tech addiction case. The executive, Adam Mosseri, 43, described Instagram, which is owned by Meta, as a steward of strong safety protocols for teenagers.

FTC Investigating Apple's Promotion of Left-Wing News Outlets

The U.S. Federal Trade Commission has raised concerns with Apple about accusations that Apple News promotes news articles from left-wing news outlets and suppresses content from conservative publications, the FTC's chairman said. In a letter sent to Apple CEO Tim Cook, FTC Chairman Andrew Ferguson noted the commission does not have authority to require Apple or any company to take positions on any political issue or to curate news based on ideology.

Subscribe

The GigaLaw Firm helps companies of all sizes protect their brands online, using domain name dispute policies – such as the Uniform Domain Name Dispute Resolution Policy (UDRP) – and other legal tools available to copyright and trademark owners on the Internet.