AI · អត្ថបទ
Data Leaked Even When All AI Agents Followed the Rules, IBM Warns
IBM ·
មើលត្រួសៗ
IBM says many companies using AI agents found that data leaked even when every component followed its rules. Each part worked as designed, but the overall system broke the intended access model. The company's 2026 breach cost report also found AI-linked breaches were rising.
ចំណុចសំខាន់ៗ
- An IBM Consulting project for a drug company found that data access rules could change when data was copied or combined across platforms.
- When AI assistants connect to databases through tools like MCP, the user's permissions and the server's permissions can differ, so the AI may return data the user should not see.
- IBM's 2026 Cost of a Data Breach report found 21% of breached organizations reported an AI-related security incident, up from 13%.
- Organizations with AI-related incidents averaged $5.33 million in costs, compared with $4.7 million for other breaches.
ហេតុអ្វីវាសំខាន់
AI agents can reach sensitive data through tools without a person knowing it, so companies need to check who can access what across the whole chain.
ប្រភពដើម
모든 AI 에이전트가 규칙을 따랐는데도 데이터는 유출되었습니다
IBM · Nov 14, 00:38 KST
Read on IBM ↗AI summarized the original article. Check the original for accuracy. Copyright belongs to IBM.
IBM Quantum · ចំណាត់ថ្នាក់ និងព័ត៌មានបន្ថែម
បន្ថែមពីវិស័យនេះ
- AWS Adds Shared Cache Skill to Cut Self-Hosted LLM Cost and Latency
AWS wrote about a shared cache that lets self-hosted AI models reuse saved work across servers, aiming to cut cost and wait time. The post shows how a coding agent can help set it up. Amazon says it can cut inference cost by up to 40% and first-token delay by up to 80%.
Amazon AWS Machine Learning blog · Nov 21, 09:14 KST - Can AI Really Find Black Friday TV Deals? A BBC Test Says Not Yet
A BBC team tested four popular AI chatbots to see if they could find a good Black Friday deal on a 55-inch TV. The results were mixed, with many links broken or out of date.
BBC News · Nov 21, 09:06 KST - Anthropic Mathematician Says Claude Helped Crack the Riemann Hypothesis
A video about a claim that an Anthropic mathematician, Levent Alpöge, used Claude to find a proof of the Riemann Hypothesis in about six weeks. The creator skips the heavy math and looks at what this could mean for AI and its future. The proof is still being checked.
TheAIGRID · Nov 21, 08:45 KST - Google Brings Duplex Calling to Gemini, Letting AI Phone Businesses for You
Google has moved its Duplex phone-calling technology into the Gemini app, so Gemini can now call businesses for you. The system has changed a lot since 2018, shifting from narrow custom models to a single stack built on Gemini.
Google Research blog · Nov 21, 08:30 KST