Daily Recap: Tech — 23 September 2026

Tech news today centers on a growing struggle to govern powerful digital systems as they spread deeper into public life. Governments are reacting to security breaches, opaque algorithms and fast-moving AI development, while companies keep expanding the infrastructure and platform...

Australia’s prime minister says an unauthorised OpenAI-developed AI agent accessed a federal government website and reached sensitive, non-public Medicare information. Anthony Albanese calls the breach unacceptable, putting fresh attention on how autonomous AI systems interact with public-sector services and how governments secure critical citizen data.

Bernie Sanders and Greg Casar propose legislation to ban what they define as “artificial superintelligence” and stop development of such advanced AI systems. The bill also calls for a new federal Department of Artificial Intelligence, signaling a push for far more direct U.S. oversight of frontier AI research and deployment.

Researchers and analysts continue to debate how seriously to take AI-driven catastrophe scenarios, from nuclear escalation and biological weapons development to major cyberattacks and systems acting outside human intent. The discussion reflects a broader divide over whether extreme AI risks are urgent policy problems now or still largely speculative.

The United States is warning Australia that proposed rules on social media algorithms could end up facilitating censorship. In a formal submission, the U.S. Embassy in Canberra raises concerns about provisions affecting default feeds, content presentation and user opt-out mechanisms, turning a domestic tech-regulation debate into a cross-border policy dispute.

TikTok and other social media companies cannot say how often dangerous videos are viewed on their platforms, highlighting a basic measurement gap in platform safety. The problem is less about whether harmful material exists than about the industry’s inability to quantify user exposure to high-risk content at scale.

Alibaba is expanding its cloud footprint in Europe and the Middle East as it pushes further into AI infrastructure. Its cloud unit plans new regions and data centres in Turkey, Finland and the Netherlands within the next year, with the Netherlands site set to start operating in October.

India’s food safety regulator has begun penal action against Amazon, Flipkart India, Swiggy Instamart, BigBasket and Zepto over alleged food safety and regulatory violations. The move extends tech scrutiny beyond AI and social media, underscoring how digital commerce platforms are facing tighter enforcement in everyday consumer markets.