英国政府计划拨款1亿英镑，成立国家人工智能生物安全实验室（AI Biosecurity Lab，AIBL），用于增强应对新兴生物威胁的能力。政府称该机构将是“世界首创”，并将与领先的人工智能企业合作；OpenAI和Anthropic还将提供价值约2亿英镑的模型使用额度，供实验室调用各自的AI模型。这些额度属于模型服务资源，并非直接现金捐款。
实验室的具体职责、组织架构和选址尚未公布。英国AI事务大臣卡尼什卡·纳拉扬称，实验室将汇集顶尖科学力量，帮助英国在人工智能发展过程中防范生物风险和网络攻击，同时支持本国科学、先进芯片及AI产业发展。
这笔投资是英国政府扩大AI与网络安全防御投入的一部分。预算还将向英国AI安全研究所追加1.2亿英镑，用于继续评估先进模型的风险；该研究所于2023年成立。政府另计划投入6000万英镑建设“国家网络盾牌”，利用代理式AI加强网络防御，并拨款7000万英镑成立国家暴露面与实时漏洞中心，以识别英国关键基础设施和经济领域的网络漏洞。
相关计划出台之际，PHP编程语言近期发现严重安全漏洞，多个机构出于预防考虑暂时关闭部分网站。研究人员称，漏洞可能使使用PHP的网站面临攻击风险；截至报道时，尚无成功利用漏洞的报告。安全专家同时警告，AI可能让潜伏多年的软件缺陷更快被发现，也更容易被攻击者利用。
本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
The British Government plans to allocate Pound100 million for the establishment of the National Laboratory for Artificial Intelligence Biosafety (AI Biosecurity Lab, AIBL) to enhance capacity to respond to emerging biological threats. The Government stated that the agency would be the “first in the world” and would work with leading artificial intelligence firms; Open AI and Anthropic would also provide a model worth about Pound200 million using quota for laboratory access to their respective AI models. These quotas are model service resources and are not direct cash contributions.
The specific functions, organizational structure and location of the laboratory have not yet been announced. The British Minister for AI, Kanishka Narayan, stated that the laboratory would bring together the best scientific forces to help the United Kingdom protect against biological risks and cyberattacks in the course of artificial intelligence development, while supporting the country ' s scientific, advanced chip and AI industry.
This investment is part of the British Government's expansion of AI and Cybersecurity Defense. The budget will also provide an additional Pound120 million to the United Kingdom ' s AI Institute for Security Studies, which was established in 2023 to continue to assess the risks of advanced models. The Government also plans to invest £60 million to build a “national network shield”, strengthen cyberdefence using proxy AI and allocate £70 million to establish a national exposure and real-time gap centre to identify gaps in the UK's critical infrastructure and economic areas.
At the time of the project, serious security gaps had recently been identified in the PHP programming language, and several agencies had temporarily closed some of their websites for preventive reasons. According to researchers, loopholes could expose websites using PHPs to the risk of attack; at the time of reporting, there were no reports of successful use of loopholes. At the same time, security experts warned that AI could make software deficiencies that had been in hiding for many years more visible and more accessible to the attackers.
Auto-fetched and translated by AI Atlas. All rights belong to the original author.