2026市场之声：AI报告-北美客户对AI应用、监管与风险的跨行业洞察

Foreword As organizations worldwide accelerate enterprise-scaleAI deployment, they face a critical challenge: balancinginnovation with legal, ethical and reputational risk. What wasonce exploratory adoption has quickly become a strategicpriority, with leaders now focused on building responsibleframeworks, accelerating implementation and capturingmeaningful business value. In the absence of universalregulation, many companies across the world are designingtheir own AI governance frameworks. Yet gaps persist incritical areas such as third-party data use, liability assignmentand intellectual property ownership. To better understand how organizations arenavigating this shift, Dentons surveyed legal,business and operations leaders following ourinaugural North American Legal AI Summit, hereis a look back at what the survey data revealed.Their responses offer a view into where legal teamsstand today: what is driving adoption, where thepressure points are, and how they are preparing forthe next wave of change. This report distills thosefindings into a clear, structured view of how legalteams are adopting AI—highlighting the big-pictureobservations, business implications and the strategicconsiderations leaders should prioritize next. Leaders want AI to empower teams. However, thereis the need for balance: enough structure to guideteams, without rules that hold them back. As onerespondent put it: “We want to cover the basics, butwe don’t want to stymie innovation or scare peopleaway with too many ‘do nots.’” Executive summary – Keysurvey insights Leaders are watching the privacygap in AI adoption The impact of AI on M&A is herealready—and clients are feeling it 39% of respondents say AI is already moderately orfundamentally transforming deal flow, valuations,synergies and diligence, and 70% report at leastsome impact. This is driving immediate demand forAI-enabled M&A advisory, diligence automation, andenhanced valuation and antitrust support. A third of participants expressed concern aboutvendor use of data, chiefly about whether vendorsare using shared data for their own purposes andfor training AI, creating real exposure around dataprotection, confidentiality, and the unauthorizeduse of proprietary information for model training. AI governance is taking shape, but adoptionof standards is inconsistent IP ownership of AI outputs is fragmented With respondents divided across joint ownership (21%),user ownership (35%), developer ownership (21%) andpublic domain (16%) ownership models, our surveyresults underscore how unsettled AI intellectualproperty ownership remains and signal demandfor AI IP strategy, contracting templatesand risk allocation models. Respondents reported drawing on governmentguidance (10%), ISO/IEC 42001 (13%) and variousindustry-specific frameworks (25%) to shape internalAI practices. What organizations need, however,is practical, implementable governance—not justpolicy. This creates a clear opportunity for structuredgovernance programs, certification readinesssupport and board-level education. Contractual challenges relating to risk allocationand responsibility More than half of respondents believe that theroles, responsibilities and risk allocation associatedwith algorithmic harm or errors in AI results areonly loosely defined. This indicates that businessesare deploying AI faster than they are defining whois responsible when things go wrong, exposingthem to unclear liability, disputes with vendors andinconsistent internal decision-making. M&A and AI :The big pictureSurvey resultsFindings and insights AI’s influence on North American M&A is becoming increasingly visible.According to 70% of our respondents, AI has at least some impact on howtransactions are analyzed and negotiated. Specifically, buy-side dealmakersemphasize proprietary data, usage growth, scalability and brand as AIvaluation drivers. Survey results also show that diligence is expanding toexamine to rights to data, modelling risk and maturity of AI governance. Question: How significantly is AI influencingmergers and acquisitions today? Results: Implications for your business AI promises to transform both the business drivers ofM&A programs, as well as how M&A transactions arecarried out. An AI-driven M&A process may include: •Instant population and review of data rooms; •AI-assisted risk detection and issue spotting;•Changing standards of materiality driven by AI’sability to review an entire universe of documentsand data points;•Representations and warranties and otherinsurance products where risk is pricedand calibrated partially based on AI reviewprocesses; and•AI-facilitated negotiation designed to achievefaster consensus and remove the friction anddelay associated with the drafting process. Companies involved in M&A should leverageAI technologies to streamline processes, reducecosts and mitigate risks throughout the transactionlifecycle. Dentons’ deal teams, including AI leaders,can help you ensure M&A, i