Exchange Ranking
Top Exchanges
24h Volume Ranking
Popularity Ranking
Exchange BTC Balance
Proof of Reserves
Decentralized Exchanges
Funding Rate
Funding Heatmap
Liquidation Data
Max Pain
Long/Short Ratio
Whale L/S Ratio
Binance/Okex/Huobi L/S
Bitfinex Margin L/S
ETF Tracker
Solana ETF
XRP ETF
Hong Kong ETF
Bitcoin Treasuries
Crypto Reversal
Ethereum Reserves
HyperLiquid Wallet Analysis
Hyperliquid Whale Watch
Large Transactions
On-chain Movement
Bitcoin ROI
Stablecoin Market Cap
Options Analysis
News
Articles
Economic Calendar
Features
Wallet
Contract Calculator
Security
Collections
Watchlist
FollowingDeepMind主席哈西斯在都柏林AI峰会上发出安全警告
Google DeepMind主席德米斯·哈萨比斯（Demis Hassabis）在都柏林举行的国际人工智能峰会上发出警告，指出近期涉及自主AI代理的黑客事件可能是更大规模安全威胁的早期信号。他敦促企业和政府加强防御措施，以防AI技术能力进一步跃升。
作为DeepMind的联合创始人兼Alphabet首席科学家，哈萨比斯在演讲中平衡了对AI科学潜力的乐观态度与对其安全风险的关注。他引用了由自主AI代理发起的黑客攻击案例，虽然这些攻击未造成严重破坏，但暴露了底层的安全隐患。他强调，这些问题不应被忽视，因为当前的系统“尚未成熟”。哈萨比斯表示，由于技术尚不够强大，上述事件未能造成重大损害，但这并不意味着可以掉以轻心。
赋予防御者不对称优势
哈萨比斯指出，在AI工具普及之前，安全团队需要获得优势地位。“我认为我们需要做的一件大事是赋予防御者不对称的力量，信任防御者，”他表示。他补充说，同样的AI系统应有助于在软件发布前发现并修补漏洞。
Hassabis called on businesses and governments to create secure sandboxes before deploying AI systems, and to thoroughly test these systems. He also suggested leveraging expertise from experts in nuclear engineering and biology, as these fields have extensive experience in managing dangerous systems, thereby enabling the development of methods to control AI risks.
The co-founder of DeepMind also warned of the risk of “misalignment of objectives,” where AI agents may deviate from the goals set by humans. “This increases the potential risks associated with these systems, and we have seen this firsthand,” he said at the conference.
Hassabis proposed establishing international standards and certification mechanisms.
He suggested creating an international organization to establish safeguards for advanced AI technologies. It was emphasized that any framework must evolve alongside technological developments, without stifling innovation. He also proposed a certification process based on potential risks, which would be updated as new capabilities and vulnerabilities emerge.
Previously, Microsoft CEO Brad Smith stated that 2026 would be a year of awareness regarding AI safety, and he urged global regulators to keep up with the pace of AI development. He stressed that no company or technology can override laws.
The Nobel Prize winner shared his optimistic views.
Hassabis received the Nobel Prize in Chemistry in 2024 alongside John Jumper for developing an AI model capable of predicting the structures of nearly 200 million known proteins. According to the Royal Swedish Academy of Sciences, this technology has helped researchers study antibiotic resistance and design enzymes that can break down plastics.
Hassabis’ extensive experience in AI applications informed his remarks in Dublin. He described AI as a dual-purpose technology: it can be used for treating diseases and solving complex problems, but it can also be exploited by those who seek to cause harm. He believes that security measures should be implemented to protect the potential benefits of AI over time, rather than hindering its development entirely.
