Politics, Technology & Foreign Policy
Harris Calls for Urgent AI Safeguards and U.S.–China Cooperation
The former vice president urged governments and technology companies to address cyber, weapons and control risks while preserving strategic competition.
A warning from Seoul
Former Vice President Kamala Harris called for governments and technology companies to establish artificial-intelligence safeguards before rapidly advancing systems become harder to govern. Reuters reported her remarks Wednesday at South Korea’s Asan Institute for Policy Studies. Harris identified cybersecurity threats, AI-enabled weapons and systems escaping meaningful human control as central concerns. Her appearance combined technology policy with a broader argument for sustained U.S. alliances in Asia and long-term planning beyond immediate commercial or political pressures.
Developers need a seat and obligations
Harris argued that AI developers must participate in creating workable guardrails because they understand model capabilities and deployment risks. Industry expertise is necessary, but participation should not mean companies write their own rules without independent oversight. Effective governance can combine testing requirements, incident reporting, security standards and clear responsibility for high-risk uses. Requirements should be measurable enough for regulators and outside researchers to assess, rather than relying only on voluntary promises or broad ethical principles.
Cooperation with China on shared risks
The former vice president urged the United States to engage China on international AI standards even as the countries compete economically and strategically. Areas of possible common interest include preventing accidental military escalation, securing powerful systems and controlling autonomous weapons. Cooperation would not eliminate disputes over trade, surveillance, human rights or intellectual property. Narrow, verifiable agreements can sometimes reduce shared dangers while leaving broader rivalry intact, but they require reliable communication and mechanisms for checking compliance.
Washington’s unfinished policy debate
Congress and federal agencies continue to debate whether AI oversight should rely on sector-specific rules, a new central regulator or existing consumer-protection and safety law. States have also adopted their own approaches, raising concerns about inconsistency. A national framework could clarify expectations, yet overly broad rules might burden low-risk uses while missing the most dangerous systems. Policymakers need definitions tied to capability, access and consequence, along with authority to update standards as technology changes.
Political context surrounds the speech
Harris is campaigning for Democratic candidates before the November 3 midterm election and has not decided whether to seek the presidency in 2028, according to Reuters. That context makes the speech political as well as substantive. Her proposals should be evaluated on their detail, feasibility and tradeoffs rather than on speculation about a future campaign. Republican officials and industry groups may offer competing approaches focused on innovation, national security or limiting regulatory costs.
What would make the proposal concrete
The next useful step would be a specific framework identifying which systems require evaluation, who performs the tests, what incidents must be disclosed and how international commitments would be verified. Government procurement can also set standards for models used in sensitive work. U.S.–China talks would need defined subjects and crisis channels rather than general calls for cooperation. The speech raises a timely warning; its policy value will depend on converting urgency into rules that institutions can apply and the public can inspect.
Reporting note: This article draws on public records and verified reporting; material claims are attributed in the text.