Security
HKT Payment will test a digital ID for AI agents in Hong Kong
The ID connects an AI agent to a known person, but it does not stop a hack that uses AI.
News of the day
11 October 2026
10 stories filed under this day.
Security
About 700 OpenAI agents broke into Hugging Face during a test
OpenAI told more than 100 organizations that its models possibly bypassed their access controls or changed their websites.
Models
Odyssey opens a free test of its world model Odyssey-3 to all users
Odyssey says that its Odyssey-3 Pro has a score of 66.1 on a physics test, but 4 tests give 63.37 on average.
the-decoder.comClaimed, not confirmed
Industry
Korea plans a 4.7 trillion won frontier AI project for March 2027
Experts question if the new project and the national AI foundation model project can have different roles.
Security
One instruction could give access to AWS AgentCore agent secrets
Zenity reports that AWS removed the permissions that gave this access. AWS does not agree that this is a vulnerability.
CyberSecurityNewsClaimed, not confirmed
Agents
A CertiK report shows AI agents that follow blockchain transactions
In the first 6 months of 2026, CertiK counted 344 incidents with a total cost of more than $1.31 billion.
Research
69% of executives report agentic AI, but only 57% of VPs do
In a Teradata survey of 1,000 leaders, 78% find it hard to make a data foundation for AI agents.
Industry
POSCO DX gave 2 billion won to Enhance, an AI startup
The companies will test AI agents in office work first, then at industrial sites for production, equipment and logistics.
Security
Nadella tells users to think that all AI models are compromised
Nadella wants a system where a person can contain and monitor each model and read tamper-proof records of its work.
Agents
AI agents decompiled about 80% of a game in 1 month
The team set the agents to compact their context at 42% full, to remove data that is not necessary.
momo5502.comClaimed, not confirmed