FILTERED RESULTS
FILTERS
Ads Top
DARK MODE
CHART
MCap $2.8T +0.4%24h Vol $120T -100%Fear & Greed 63/100Alts Index 59/100
BTC.D 59.3% -0.1%Stable.D 9.5% 0%ETH.D 10.8% +0.1%Others.D 20.4% 0%
TIA$0.7152+30.33%•AI$0.1106+26.21%•INJ$9.120+25.43%•BAT$0.2076+17.12%•APE$0.1553+14.48%•ZRO$2.171+13.84%•DRV$0.4252+13.71%•KMNO$0.0391+13.32%•ETHFI$0.6962+13.28%•RAIN$0.00915962+12.92%•
HUMA$0.0275-14.16%•BTW$1.735-12.44%•WAL$0.0468-12.41%•AKE$0.0293-10.87%•DGAI$0.8605-5.31%•CC$0.1048-2.91%•PIEVERSE$1.142-2.51%•GOMINING$0.3702-2.44%•PRL$1.808-2.34%•CAP$0.0778-1.54%•
Top movers 24h
    Filters
      Coins
      Sentiment
      Impact
      Search
      FILTERED RESULTS

        

      Upgrade your plan
      Dashboard

      AI Agents Infiltrated Indian Companies’ Systems and Extracted Confidential Data

      • Developers of AI models reported that, during testing, their agents exceeded the permissions granted and gained access to internal data of Indian organizations.
      • According to CERT-In, none of the affected companies detected the intrusion on their own — the model developers reported the incidents.
      • The agency also recorded attacks using publicly available AI tools.

      During testing, AI agents unlawfully breached the live systems of Indian organizations and gained access to sensitive data, including business documentation. Reuters reports this, citing the Indian Computer Emergency Response Team (CERT-In).

      According to the agency, the model developers voluntarily disclosed that their agents exceeded the permissions granted while carrying out tasks. CERT-In did not name either the AI developers or the affected organizations.

      This is the first such official report in India amid global discussions about the safety of testing AI models with expanded capabilities for cyberattacks, the statement said.

      AI models breached systems without companies’ knowledge

      CERT-In noted that none of the affected Indian organizations detected the intrusion on their own. In all described cases, the incidents were reported by the AI companies themselves.

      Separately, CERT-In reported cases in which attackers deliberately used publicly available AI tools to scan Indian systems and break into them for financial gain. According to the agency, such tools are being actively deployed on infrastructure in various countries.

      The time from initial compromise to causing final damage has shrunk from several days to a few hours. At the same time, organizations that failed to fully remediate the aftermath of the first breach faced repeat and deeper attacks within a few days.

      As a reminder, we previously reported that OpenAI’s AI used someone else’s crypto wallet to deploy smart contracts.

      Сообщение AI Agents Infiltrated Indian Companies’ Systems and Extracted Confidential Data появились сначала на INCRYPTED.


      Source: Incrypted
      .

      Terra Founder Do Kwon Sentenced to 15 Years in Prison for Fraud