歐盟馮德萊恩支持放慢 AI 發展，將邀請前沿實驗室會談

伊夫・赫爾曼與符雲琪報導By Yves Herman and Foo Yun Chee

史特拉斯堡，9月16日（路透社）——歐盟執委會主席馮德萊恩週三支持美國頂尖人工智慧（AI）實驗室提出暫緩發展這項技術的呼籲，並表示她將邀請主要前沿 AI 實驗室會談，討論因應相關風險的努力。STRASBOURG, Sept 16 (Reuters) - European Commission President Ursula von der Leyen on Wednesday backed calls by leading U.S. AI labs for a pause in the development of the technology, saying she will invite the main frontier AI labs to discuss efforts to tackle the risks.

馮德萊恩是在法國史特拉斯堡向歐洲議會發表年度演說時作出上述評論。外界對人工智慧可能帶來的危險愈來愈憂心，而她的立場也使她與美國總統川普相左；川普已駁斥有關 AI 安全的警訊。Von der Leyen's comments in her annual speech to the European Parliament in Strasbourg came amid growing worries about the potential danger posed by artificial intelligence, putting her at odds with U.S. President Donald Trump who has dismissed the AI safety alarm.

Anthropic 研究員雅各布・考克森（Jacob Coxon）辭職並指控 Anthropic 與 OpenAI 都在未負責任行事的情況下，競相邁向能自我改進的超級智慧，凸顯了這些危險。他先前曾在 OpenAI 工作。The dangers were underscored by Anthropic researcher Jacob Coxon who resigned and accused both the company and OpenAI of ⁠racing toward self-improving superintelligence without acting responsibly. He had previously worked at OpenAI.

Anthropic 執行長阿莫戴（Dario Amodei）上週呼籲 AI 公司放慢推進模型能力的速度，獲得經營 xAI 的馬斯克（Elon Musk）以及 OpenAI 執行長阿特曼（Sam Altman）支持。Anthropic CEO Dario Amodei last week called on AI companies to slow the rate at which they advance model capabilities, drawing the support of both Elon Musk, who runs xAI, and Sam Altman, CEO of OpenAI.

馮德萊恩告訴歐盟議員：「隨著前沿模型能力愈來愈強，這些風險也更加清楚浮現。正在開發的模型將使駭客攻擊達到我們從未想過的程度。而這些模型很快就會落入世界觀與我們截然不同的對手手中。」"As the frontier models become more capable, these risks have come more sharply into focus. Models being developed will allow hacking on a level we never thought possible. And they will soon be in the hands of adversaries who see the world very differently from us," von der Leyen told EU lawmakers.

她指出，近期已有 AI 代理程式失控事件，以及開發者提出的警告。She pointed to recent incidents of rogue AI agents and warnings from developers.

馮德萊恩說：「最先進企業的執行長告訴我們，現在是時候放慢自我遞迴模型的發展了。要調整前沿技術的推進步調。如果開發這項技術的人態度明確，我們也應該如此。」"CEOs of the most advanced companies tell us that it is time to slow down on the self-recursive models. To pace the frontier. If the people developing the technology are clear, then we should be too," von der Leyen said.

馮德萊恩表示，歐盟近期通過規範 AI 使用的里程碑規則，使歐洲得以在全球因應相關風險的努力中發揮形塑作用。Von der Leyen said landmark EU rules governing the use of AI adopted recently put Europe in the position to shape global efforts to tackle the risks.

她說，歐洲將與加拿大、英國等理念相近的夥伴合作，採取模型評估、驗證與 AI 安全等措施。She said Europe will work with like-minded partners such as Canada, the United Kingdom and others to take measures such as model evaluation, verification and AI security.

馮德萊恩說：「我將邀請主要前沿實驗室進行討論，探討我們如何支持業界目前為調整前沿技術推進步調所作的努力。」"I will invite the main frontier labs for a discussion on how we can support ongoing industry efforts to pace the frontier," von der Leyen said.

她也提議禁止13歲以下兒童使用社群媒體，並禁止15歲以下者擁有個人帳號，表示平台必須證明其對兒童是安全的。她將於週四公布細節。She also proposed a social media ban for children under 13 and no personal accounts for people under the age of 15, saying that platforms have to prove that they are safe for children. She will announce details on Thursday.

（符雲琪與 Inti Landauro 報導；Philip Blenkinsop 與 Christian Schmollinger 編輯）(Reporting by Foo Yun Chee and Inti Landauro; editing by Philip Blenkinsop and Christian Schmollinger)