Gartner 推出 Value-Risk Flywheel 框架,帮 CIO 判断 AI 成果
Gartner Introduces Framework for CIOs to Determine AI Outcomes
Gartner 发布面向 CIO 的 Value-Risk Flywheel 框架,用“动量”和“阻力”两种力量评估 AI 成果,主张把 AI 价值与风险对齐。Gartner 调查显示,71% 的组织计划在未来 12 个月内部署 agentic AI,但仅 21% 具备 agentic AI 治理机制。该框架强调需要中央决策机构判断价值何时值得承担风险。
正文
Gartner has released the Value-Risk Flywheel for CIOs. According to the company, the framework is meant to align the value AI creates for enterprises with risks.
“The best thing a CIO can do is to align value with risk and align safety with governance,” Daryl Plummer, Distinguished VP, Managing VP, and Chief of Research at Gartner, said in a statement.
The Flywheel framework assesses momentum and drag as the two forces shaping AI outcomes. Counterintuitively, momentum isn't always good, and there is certain drag that is beneficial, per Gartner.
“CIOs who outperform know how to create positive momentum and strategically introduce good drag, while reducing bad momentum and bad drag,” Plummer added.
For instance, in software development, value can compound faster than risk since AI can create a significant share of code, which relieves software developers from the drag of manual coding and shifts their focus to rapid prototyping, requirements-driven development, architectural thinking, and product engineering.
“Flywheels need a governor, right at the centre, who decides when value is worth the risk, what perspective on value should dominate and how to keep an organisation safe,” he noted.
“It is a centralised decision authority that drives advice and guides actions on accountability, technology and economics,” remarked Kristin Moyer, Distinguished VP Analyst and Gartner Fellow.
The Flywheels initiative comes in as a Q2 2026 Gartner survey of over 2,000 CIOs and technology executives found 71% of organisations expect to deploy agentic AI within the next 12 months, while only 21% have agentic AI governance mechanisms. According to Moyer, this imbalance complicates creating value quicker than AI is increasing risks.