全球各地的企业内部法律团队正积极探索人工智慧的创新应用方式，以加速工作流程、降低风险并腾出时间处理更有价值的专案。日本金融科技公司Upsider的首席法务官自行开发AI工具，将Slack和Notion上的对话结合内部法律手册输入Anthropic的Claude模型，自动标记潜在的监管与合约风险，使法律团队能在业务部门寻求协助之前主动介入。美国Cox Media Group则将合约处理工具改造为政治广告审查系统，以应对诽谤风险，而新加坡南洋理工大学则利用Lexagle合约平台强化对不当行为举报的调查，将案件平均结案时间从55天缩短至38.5天。
科技服务公司Endava运用OpenAI的ChatGPT打造客制化工具，处理保密协议评估及第三方筛查等痛点，第三方筛查时间减少了40%至50%。该公司定期举办黑客松活动并邀请OpenAI代表参与，促进团队协作与技术适应。在人力方面，多家公司展现出以精简团队应对业务增长的趋势：Endava法务团队维持约35人的稳定编制，而Upsider在营收以每年约50%速度增长的情况下，仅计划将内部律师从两人扩充至四人。
尽管企业法务部门的AI采用率在一年内翻倍，且近三分之二的团队预期将减少对外部律师的依赖，但预算增长并未跟上脚步。企业法务营运联盟Cloc执行长指出，法律团队被要求「以更少资源做更多事」，因此倾向于让单一AI工具实现多重用途以提高效益，但这也带来验证不足的潜在风险。Gartner分析师将当前状态比喻为「自助餐」——法律团队正在广泛尝试各种AI方案，从人力资源平台到合约与第三方风险工具的整合，目前尚不确定哪种方案能带来最大回报。（关键数字： a year）
In-house legal teams worldwide are creatively adopting AI to accelerate workflows, reduce risk, and free up capacity for higher-value work. Japan-based fintech Upsider built a custom tool feeding Slack and Notion conversations into Anthropic's Claude alongside an internal legal playbook to flag regulatory and contractual red flags proactively. Cox Media Group in the US repurposed a contract-processing tool for political ad reviews to mitigate defamation risks, while Nanyang Technological University in Singapore leveraged Lexagle's contract platform to strengthen misconduct investigations, cutting average case closure time from 55 days to 38.5 days.
Tech services company Endava has used OpenAI's ChatGPT to build customised tools addressing pain points such as NDA evaluation and third-party screening, achieving a 40 to 50 percent reduction in screening time. Regular hackathons with OpenAI representatives foster collaboration and help the company adapt as AI evolves. Several organisations are demonstrating that lean teams can handle growing workloads: Endava's legal team remains stable at around 35 people, while Upsider plans to expand from just two to four in-house lawyers even as revenue grows at roughly 50 percent annually.
Although AI adoption among in-house legal teams has more than doubled in a year and nearly two-thirds expect to reduce reliance on outside counsel, budgets are not keeping pace. Cloc's CEO notes that legal departments are expected to do more with less, driving them to seek multiple uses from single AI tools — though this increases inherent risk when validation for one use case does not extend to another. A Gartner analyst compares the current landscape to a buffet: legal teams are broadly experimenting with options ranging from HR compliance platforms to integrated contract and third-party risk tools, uncertain which will deliver the most promising payoff.