(This article is part of Today’s Cache, The Hindu’s newsletter on emerging themes at the intersection of technology, innovation and policy. To get it in your inbox, subscribe here.)

Meta defeats child safety lawsuit

Meta and CEO Mark Zuckerberg has won a shareholder lawsuit that claimed they were misled about child safety on Facebook and Instagram. The judge finally ruled yesterday that there was insufficient proof that the shareholders had dealt with financial losses due to Meta’s alleged disclosures. Meta will also not need to reveal how severe the sexually explicit content is on its platforms or any plans for child protection that it may have considered. 

The lawsuit intended to block Meta from holding their 2024 annual meeting until the proxy statement was amended and void election results if the meeting was held. Despite the dismissal of this case, Meta is fighting multiple lawsuits alleging it encouraged children to get addicted to its platform while downplaying the risks. 

TSMC told U.S. of chip in Huawei product

Taiwan Semiconductor Manufacturing Company told the U.S. about one of its chips in a Huawei’s latest AI processors, Ascend 910B after a report was published by tech research firm TechInsights. This could mean that TSMC was possibly violating export restrictions placed on the Chinese company by the U.S. 

The tech research firm took apart the AI processor to find TSMC’s chip among others in a multi-chip system. The 910B is the most advanced AI chip available in the market from a Chinese company. TSMC notified the U.S. Commerce Department saying they hadn’t supplied chips to Huawei since mid-September 2020. The U.S. department acknowledged the violation made by TSMC but didn’t confirm if the company would be investigated further. 

Anthropic releases autonomous AI for coders

Alphabet and Amazon-backed AI startup Anthropic has released new AI models with the ability to autonomously perform computer tasks and save users keystrokes. The new feature called, “computer use” is able to direct where the mouse needs to move, where to click and what to type for some complex tasks, the company said in a demo. 

The tool is essentially an AI agent which are able to navigate through apps and perform a full task completely on its own without any human intervention. AI agents have recently gained popularity and are said to be the future after AI chatbots. The new updates are for Anthropic’s mid-tier AI model, Sonnet and its cheapest model, Haiku.