并购聚焦：人工智能与并购战略

M&A Spotlight:AI and M&A strategy 26 February 2026 Luke Templeman | Thematic Strategist |Luke.Templeman@db.com| +44 207 541 0130Galina Pozdnyakova | Research Analyst |Galina.Pozdnyakova@db.com| +44 207 547 4994 The recent equity market volatility and AI-related sell-off has sent CEOs scrambling to makesure they have an AI strategy and can articulate it to their investors Software has predictably been the worst performing industry while the megatrend of energy has boosted its appeal. The US sell-off hasbeen more dramatic than that seen in Europe, however the same picture emerges when we look across the industries in the Stoxx600. Yet, only 11% of companies were likely to have at least one AI-related business function fullyimplemented in 2025. That leaves most CEOs under heavy pressure to speed their AI response Where is your organisation currently in theimplementation of agentic AI?(Survey in Q2 2025) How much work can AI do? The IMF last year highlighted the pros and cons of using AI in the workplace. Itargued that about 40% of global employment will be affected by AI, particularly“cognitive” work. However, it also pointed to a recent study that showed that, insome roles, although the overall quality of output may be boosted by AI, thetechnology as it stands can also produce more “sameness” given that the toolsused give something of a standardised response. Startups and large businesses are earlyadopters. Mid-sized firms may be left behindProjected AI usage rate over the next 6m by business size Hence, we see corporate M&A involving AI as being critical to the adaptation plans of manyCEOs, especially those who need to catch up to their peers Software M&A by non-tech corporates has room to grow, especially as companies consider buying v buildingcertain capabilities. Such M&A’s share of global software dealmaking (by value) has shrunk from 17% in 2000s to5% in 2020s, with tech buyers (45% to 23%) also sidelined amid a rise in sponsored and similar transactions (28%to 72%). Those big private equity deals would need an exit, and a sale to corporate will be among the key routes. As valuations have dropped, the key question for M&A markets is whether multiples for unlistedtech targets have also repriced as acutely as they have in listed markets Software & services used to be the 3rdmost expensive industry group, it’s now 9th(3rdto 13thin Europe), down by 5.8x (5x). No otherindustry group has seen a similar multiple re-rating. Adjusting for growth expectations, the valuation of software companies is now fairly average When we look at PEG ratios (which adjust for growth expectations) although data is less available, software & services in theUSslidfrom 7thto 17thmost expensive industry group (3rdto 15thin Europe) Is AI booming or busting? We believe AI will likely continue to power a market rally into themedium term, at least. Experience from the 1990s shows that a long tech rally can continue despite many scares along the way. The Nasdaq surged +32%p.a.in the 1990s despite numerous >10% drawdowns. AI and M&A strategy: Questions for CEOs Business model questions •Investor expectation managementInvestor expectation management •Have you analysed how AI will affect your industry’s products,costs, and supply chains (including inputs, processes, paymentsystems, and end market offering)?•What new AI competencies do you need?•How strong is your moat from AI disruption (e.g. IP, unique data,exclusive contracts etc)?•Is your current AI-related R&D having a tangible businessimpact?•Does your business model need to pivot? •What timeline are your investors pushing you towards?•What are the key non-financial metrics to consider?•How is AI championed at a board level?•What is the target’s IP protection strategy?•How will you justify the valuation of the acquisition target? •M&A questionsM&A questions Post-integration •What are the measures for success with the acquisition?(cost/revenue synergy, process slimming, financial)•Can the new AI infrastructure handle existing data volumes andfuture growth?•How to retain key talent in the target•What regulatory issues may arise?•What is the ‘burn rate’ of the target as a business unit? Whatfuture capex andopexis required?•What are the public relations risks? •Are your peers considering acquisitions?•Should youacqui-novate or buy an established business?•Do you need to buy a moat?•Are you buying AI to leverage your existing business or tovertically integrate•What is the cost delta between buying and buildingnecessary AI infrastructure? Our M&A forecasts for the US show a holding pattern Our M&A indicator for the US shows the pickup forecast for Q1 may slow going into Q2. That comes amidst rising policy uncertaintyand mixed capital issuance signals. Strong markets up to February have been a positive, so a sustained turn in sentiment maydrag onfuture dealmaking. Our M&A forecasts for the Eurozone and UK lead us to predict some short-term weakness Higher soverei