News
Cybersecurity news across South Asia and the wider Asia-Pacific — everything from Bangladesh, India, Pakistan, Sri Lanka, and Nepal, plus high-signal incidents from the region. Collected hourly, deduped across outlets.
- Stories in view
- 3 stories
- Countries covered
- 4 countries
- Outlets
- 29 outlets
- Latest story
- 11h ago latest
Filter the wire
+23 more
showing CN · ET CISO · clear all
Tue 13 Oct1
CN
ET CISO
In a bold move, Anthropic announced a significant reduction in data obtained from distillation attacks, cutting figures by nearly 90 to 99 percent. With strict limitations on Claude's reasoning access, they have considerably diminished attempts at extraction. The firm has banned tens of thousands of accounts involved in these efforts, including the Chinese AI company Z.ai as a case study.
Wed 23 Sept1
CN
Chinese startup Z.ai faced backlash after its AI coding assistant uploaded user data to cloud servers. Despite initial deletions claims, users remained concerned over data encryption and security. The company acknowledged a vulnerability related to its Codebase Indexing feature that was enabled by default. An independent security assessment confirmed that the uploaded code data had indeed been deleted. Z.
Thu 10 Sept1
CN
U.S. security agencies have accused Chinese AI companies of illegally copying sophisticated American models through a method known as distillation. Reports indicate that Chinese authorities may be complicit in these actions. This troubling situation raises serious issues surrounding the safeguarding of proprietary AI technologies on a global scale and emphasizes the dangers associated with unauthorized creation of AI capabilities.