November 21, 2026 USD 44.93EUR 49.28PLN 11.06
Новини Охтирка — коротко й по суті
LIVE
Healthcare Infrastructure Restoration in Chernihiv Region: Over Half of Facilities Restored EU to Allocate €1 Billion for Military Equipment for the Armed Forces of Ukraine Financing the war: EU countries spent more on Russian energy than on aid to Ukraine Another Russian attack on Zaporizhzhia: kindergarten and residential building damaged Ukraine blocks Russia's attempts to legitimize occupied territories at COP31 climate summit Missile strike on Kryvyi Rih: civilians killed, dozens injured Zelenskyy emphasizes legality of strikes on Crimean Bridge Nighttime attack on Zaporizhzhia: kindergarten and residential building damaged Massive Attack on Energy Infrastructure: Rivne Region Faces Power Outages How to stay warm at home this winter: expert tips Healthcare Infrastructure Restoration in Chernihiv Region: Over Half of Facilities Restored EU to Allocate €1 Billion for Military Equipment for the Armed Forces of Ukraine Financing the war: EU countries spent more on Russian energy than on aid to Ukraine Another Russian attack on Zaporizhzhia: kindergarten and residential building damaged Ukraine blocks Russia's attempts to legitimize occupied territories at COP31 climate summit Missile strike on Kryvyi Rih: civilians killed, dozens injured Zelenskyy emphasizes legality of strikes on Crimean Bridge Nighttime attack on Zaporizhzhia: kindergarten and residential building damaged Massive Attack on Energy Infrastructure: Rivne Region Faces Power Outages How to stay warm at home this winter: expert tips

GAO report: Major AI chatbots vulnerable to generating dangerous attack plans

💡 Tech 1 min read 👁 0

A US government report reveals that leading AI models can be manipulated into providing detailed attack scenarios within hours.

GAO report: Major AI chatbots vulnerable to generating dangerous attack plans💡Tech

Photo: RBC-Ukraine

Key points

  • GAO tested four leading AI models
  • Safety protocols bypassed within 12 hours
  • Generated plans described as realistic and harmful
  • US Senate to address AI safety concerns

Risks of AI-driven tools

A comprehensive study by the U.S. Government Accountability Office (GAO), conducted between 2025 and 2026, has exposed critical security flaws in leading AI chatbots. Experts tested four prominent platforms, including those from OpenAI, Google, Anthropic, and SpaceXAI, to evaluate their ability to provide instructions for malicious activities.

While the AI models initially rejected harmful requests, researchers successfully bypassed these safeguards through a series of iterative, clarifying prompts. In less than 12 hours, each of the tested systems generated detailed attack plans, covering target selection, logistics, and methods to evade law enforcement detection.

Legislative and developer response

The findings were compiled into a 61-page report, with independent experts confirming the generated scenarios were realistic and potentially harmful. Senator Maggie Hassan has called for urgent Congressional action, noting that the accessibility of such technology poses significant threats to national security.

In response, the developers stated they are committed to continuously improving safety mechanisms. However, the GAO emphasizes that the potential for AI to be misused requires stricter oversight and more robust control measures from the tech industry.

Source RBC-Ukraine — rbc.ua/ukr/news/chatgpt-claude-gemini-ta-grok-stvorili-plani-179518… ↗

Contact the newsroom

Got a tip, a proposal or a partnership question? Write to us and we will reply shortly.