Article AI

AI Today: Meta's Data Practices Under Scrutiny, Energy Demands Rise

AI News Digest

This week's AI developments highlight growing tensions between technology companies and the public, creators, and environmental advocates.

Meta Scales Back AI Image Tool Amid Criticism

Meta has scaled back its AI image training tool after facing criticism for automatically accessing public Instagram images without explicit consent from users.

AI Companies Lobby Australia on Copyright Laws

Major AI companies are lobbying Australia to weaken copyright protections for training data, prompting strong opposition from artists and exposing divisions within the governing Labor party.

Energy Demands for AI Infrastructure Spark Debate

The rapid expansion of AI infrastructure is driving a significant increase in data center energy consumption, prompting utilities and tech companies to consider natural gas as a bridge power source while renewable energy advocates push for cleaner alternatives.

Lightweight AI Model Runs on Minimal Hardware

Researchers have developed Colibrì, a proof-of-concept system capable of running a 1.5-terabyte AI model using only 25GB of RAM, potentially enabling more accessible local AI deployments.

South Korea Flags Potential AI Involvement in Recent Banking System Intrusions

South Korea's financial authorities have raised concerns that threat actors may have employed artificial intelligence capabilities in a series of cyber intrusions targeting the nation's banking infrastructure. The warning from the country's financial regulator underscores a growing trend where AI is increasingly being incorporated

Utah Tests AI-Only Patient Screening for Routine Prescriptions

A limited pilot program in Utah is testing whether artificial intelligence can handle routine patient examinations and prescription renewals without direct physician involvement. The initiative targets specific, low-complexity patient scenarios — such as routine prescription refills or straightforward symptom assessments — where the clinical question is well-defined and the risk profile is

Denmark's Historic Data Breach: 8 Million Records Stolen in Government Cyberattack

A Breach of Unprecedented Scale Denmark's government has confirmed a cybersecurity breach of historic proportions, with hackers stealing personal records of approximately 8 million citizens from a government database. The stolen data includes names, addresses, and state-issued identification numbers—a combination that could facilitate identity theft and sophisticated

AI News Digest

Meta Scales Back AI Image Tool Amid Criticism

AI Companies Lobby Australia on Copyright Laws

Energy Demands for AI Infrastructure Spark Debate

Lightweight AI Model Runs on Minimal Hardware

Read next