AI Will Negotiate Your Customer’s Next Car Deal

Your CRM Still Can’t Import an E-mail Audience. Two futures for automotive retail, and why the dealer technology stack is not ready for either one.

A strategic guide for dealer principals. Originally created March 3, 2026.

TL;DR

• Two dueling reports, Citrini's AI-crisis scenario and Citadel's data-driven rebuttal, sketch very different futures. The dealer tech stack is not ready for either one.
• The dealership model earns on friction: negotiation spread, F&I at roughly $3,000 per vehicle, service capture. Consumer AI agents remove friction and information asymmetry.
• The infrastructure underneath is the real problem: a DMS duopoly with a $129.5 million antitrust settlement behind it, CRMs that fail at basic tasks, OEM subsidies that distort buying decisions, and a vendor fragmentation tax.
• Part IX is the playbook: a survival checklist for dealer principals, plus a timeline of what happens when.

Executive Summary

In February 2026, two documents collided in the financial markets and sent shockwaves through every industry that depends on human intelligence for its revenue model. The first, published by Citrini Research on Substack, painted a scenario from the fictional vantage point of June 2028 in which artificial intelligence had triggered a cascading economic crisis: unemployment at 10.2%, the S&P 500 down 38%, and the $13 trillion mortgage market cracking under the weight of displaced white-collar workers. The second, published days later by Citadel Securities, pushed back with a data-driven rebuttal: current labor data shows no imminent displacement, technology adoption follows an S-curve, and physical constraints on compute prevent the kind of runaway automation Citrini describes.

IBM dropped 13%. Michael Burry amplified the Citrini piece on X. The debate went from niche financial analysis to front-page economic conversation overnight.

But here is what neither report addresses directly, and what this document exists to confront: What do these two scenarios mean for the 16,700 franchise automobile dealerships in the United States? What happens to the business model that has survived for over a century when AI agents can negotiate prices, compare F&I products across twenty lenders in seconds, schedule their own service appointments, and potentially bypass the dealership altogether?

The uncomfortable truth is this: your technology stack is not ready for either scenario. You are running on legacy dealer management systems controlled by a duopoly found liable in a $129.5 million antitrust settlement. Your CRM charges five figures a month but cannot perform basic functions that free consumer tools handle effortlessly. Your website platform may be running years behind on critical software updates and security patches. You are paying tens of thousands of dollars a month for tools that treat the ability to communicate with your own customers as a premium add-on.

Meanwhile, every one of these vendors is bolting on AI features, not to liberate your data or improve your operations, but to lock you deeper into their walled ecosystem. And underneath it all, the OEMs are subsidizing many of these tools, distorting the market in ways that suppress innovation and prevent you from making rational purchasing decisions.

This document is written by someone who lives this reality every day. It is written for the dealer principal who has been told to "just use the tools you have" while watching those tools fail at the most elementary tasks.

PART I: THE TWO SCENARIOS — WHAT THE REPORTS ACTUALLY SAY

Scenario A: The Citrini Thesis — Abundant Intelligence Destroys Friction-Based Business Models

Citrini Research, led by James Van Geelen and co-authored with Alap Shah (a former Citadel analyst turned AI entrepreneur), constructed a detailed, internally consistent scenario in which AI capability compounds so rapidly that by mid-2027, the economic consequences become impossible to ignore. Their thesis rests on a chain of events worth examining because each link has a direct analog in automotive retail.

The scenario begins with a step-function improvement in agentic coding tools in late 2025. By Q2 2026, procurement teams at Fortune 500 companies are watching their internal AI teams replicate mid-market SaaS products in weeks. The long tail of enterprise software begins to collapse.

The truly devastating mechanism is what happens when AI agents begin to operate on behalf of consumers. By Q4 2026, every major AI assistant has integrated agentic commerce. These agents run in the background, optimizing every consumer transaction.

The economy is built on friction, and AI removes friction. Much of what companies charge for is based on things being slow, complicated, or requiring human coordination. If AI makes those things fast and cheap, entire business models collapse.

For automotive retail, the implications are existential. If AI agents can negotiate vehicle prices by comparing every available unit within a 500-mile radius in seconds, the negotiation advantage dealerships have historically held vanishes. If those same agents can compare F&I products across dozens of lenders, the most profitable department in the dealership faces margin compression that no amount of sales training can overcome.

But here is the part Citrini does not mention, because they are not dealers: the AI agent does not even need to be particularly sophisticated to disrupt our F&I process. It just needs to do a Google search. The information asymmetry we rely on is held together with duct tape and the customer's exhaustion after four hours in the showroom.

Scenario B: The Citadel Thesis — Historical Patterns Hold, and AI Becomes a Complement

Citadel Securities, in a research note authored by macro strategist Frank Flight, offers a fundamentally different reading. Their argument is grounded in current data, not projection. Daily use of generative AI for work has been remarkably stable through 2025. No inflection point. No imminent mass displacement.

Citadel makes several economic arguments relevant to us. First, displacing white-collar work at Citrini's pace would require orders of magnitude more compute intensity than current utilization. Second, AI is a productivity shock, like steam power or the personal computer. Every major productivity shock has expanded total economic demand rather than shrinking it. Third, software engineering postings are up 11% year over year. New business formation is at record levels. These are not indicators of an economy approaching a displacement crisis.

For automotive retail, the Citadel scenario suggests AI changes the composition of dealership work rather than eliminates it. Customer-facing roles shift toward experience management. Back-office operations become more efficient. The dealership of 2030 looks different but still exists, still employs people, and still makes money.

I will be honest: the Citadel scenario is more comfortable, and that is precisely why it is more dangerous. Comfort breeds complacency, and complacency in this industry usually means signing another five-year DMS contract because switching sounds hard.

Where Both Reports Agree — And Why That Matters Most

The most important insight comes from the areas of agreement:

• AI will change the composition of work. Neither report argues nothing will happen. The debate is about speed, not direction.
• The AI capital expenditure cycle is massive. At $650 billion in 2026 alone, the infrastructure buildout is real.
• Compute and energy are genuine constraints. Physical reality imposes limits on deployment pace.
• Friction-based business models are vulnerable. Citadel disputes only the timeline, not the direction.
• Institutional response speed is the critical variable. Both reports agree: how fast you adapt determines whether disruption is manageable or fatal.

The question is not whether AI will change the dealership model. It is whether your dealership can adapt fast enough. And the answer depends on something neither report discusses: the technology infrastructure you are actually running on.

PART II: THE FRANCHISE DEALERSHIP — AN INDUSTRY BUILT ON FRICTION

The Anatomy of Your Revenue Model

To understand why these scenarios matter, look at where the money comes from. The franchise dealership revenue model is a study in friction monetization.

Your two highest-margin profit centers are vulnerable in fundamentally different ways. F&I is vulnerable to information disintermediation, which AI can deliver almost immediately. Fixed operations is vulnerable to physical disintermediation, which takes much longer. Presidio's February 2026 report showing dealership profit falling 9.7% in Q4 2025, despite gains in F&I and fixed ops, should alarm you. If profits are falling when your two best departments are performing well, the structural economics are deteriorating underneath you.

The Franchise Protection Framework — A Legal Moat Under Siege

Franchise dealerships have survived for over a century partly because every state has dealer protection laws. That legal moat is being tested. Scout Motors obtained direct-to-consumer licenses in Colorado in December 2025. Washington State is fast-tracking legislation to enable new EV brands to sell directly. The Georgia Supreme Court is examining constitutional limits on direct-sales bans.

But the real threat is not from manufacturers going direct. It is from AI agents making your intermediary role less valuable. Franchise laws protect your right to be the point of sale. They do not protect your ability to extract margin from the transaction. No franchise law can defend against that.

PART III: THE DMS PROBLEM — HOW LEGACY INFRASTRUCTURE HOLDS THE INDUSTRY BACK

The Duopoly

The DMS is the operating system of your dealership. Every transaction, every customer record, every accounting entry flows through it. And this critical infrastructure is controlled overwhelmingly by two companies whose combined antitrust settlement of $129.5 million, approved in February 2025, covered more than a decade of anticompetitive behavior. That settlement is public record and should concern every dealer principal.

The technical architecture of legacy DMS platforms is, by design, a closed system. Third-party access is restricted through prohibitive integration fees and contractual provisions. Developers receive licenses described as limited and revocable. Dealerships spend an average of approximately $42,000 per year just accessing their own data across systems. Innovation is throttled at the DMS layer. Every new tool, every AI application, every improvement that requires DMS data must pay the toll or operate blind.

The Cyberattack That Proved the Fragility

In June 2024, a ransomware attack took down the largest DMS provider's entire SaaS platform, rendering approximately 15,000 dealerships inoperable. Employees tracked orders on sticky notes. Sales teams could not process financing. Service departments could not look up repair histories. Total economic damage exceeded $1 billion. New vehicle sales declined 7.2% in June.

The most revealing aspect was the response: there was no response. Dealerships had no fallback systems. No alternative data access. They stopped functioning entirely. Dealers who ran both their DMS and CRM on the same parent company's products lost everything simultaneously. It was the single largest proof that vendor concentration is an existential risk.

PART IV: THE TECHNOLOGY STACK CRISIS — AN INDUSTRY RUNNING ON YESTERDAY'S TOOLS

The CRM Gap

If the DMS is the operating system, the CRM is supposed to be the brain. It should track every customer interaction, automate follow-up, manage your sales pipeline, drive marketing campaigns, and give you a unified view of customer relationships. In 2026, with automotive CRM platforms costing dealers anywhere from $5,000 to $15,000 per month, you would expect these systems to perform at least as well as the consumer-grade tools available to a college student running an Etsy shop.

They do not. As a category, automotive CRMs share a set of structural deficiencies that would be comical if they were not costing you money:

• Closed ecosystems. Most cannot integrate with standard marketing automation tools that power the rest of the economy. Platforms with millions of users and hundreds of integrations are locked out because the automotive CRM treats every outside tool as a competitor.
• Basic communication failures. Features that every free email marketing tool provides, such as importing a contact list or sending a campaign to your customer base, are either unavailable or require expensive add-ons.
• No native advertising integration. In a world where attribution is everything, automotive CRMs cannot natively connect to major advertising platforms. Dealers rely on third-party middleware for connectivity that mainstream CRMs provide out of the box.
• AI bolted onto broken foundations. Vendors are racing to add AI features, but these AI tools are built within the same walled gardens. The AI cannot detect when a human salesperson is already in conversation with a customer. It cannot respect business hours. It penalizes salespeople when customers exercise their legal right to opt out of messages. These are not edge cases. They are fundamental design failures.
• Missing service modules. In an industry where fixed operations is the most defensible profit center, many automotive CRMs lack robust service functionality.

Compare this to what the rest of the business world uses. Mainstream CRM and marketing platforms offer thousands of integrations, native advertising connectivity, open APIs, advanced automation, robust service modules, and mobile experiences that actually work. They cost a fraction of what automotive CRMs charge. The only reason they do not dominate automotive retail is that they lack native DMS integration. The walled garden protects the incumbents. It does not protect you.

The Website Problem

Your dealership website is your digital storefront. It is the first impression for 95% of car buyers who begin research online. According to a September 2025 study by Overfuel, 99.6% of dealer websites from the top 50 dealership groups in North America fail Google's Core Web Vitals assessment. Out of 1,910 retail websites studied, only seven passed on both mobile and desktop. The study found that failing Core Web Vitals costs dealerships approximately $30 of every $100 in advertising spend through inflated ad costs, weaker quality scores, and reduced organic visibility.

The problems are systemic across the major dealer website platforms:

• Outdated infrastructure. Some platforms are running content management systems and SEO plugins that are years and multiple major versions behind current stable releases. This is not a minor version gap. It means missing years of performance improvements, security patches, and capabilities that competitors outside automotive take for granted.
• Template duplication. SEO content, including meta descriptions, page titles, and landing page copy, is duplicated across hundreds or thousands of dealer sites. Search engines penalize this, and your organic rankings suffer.
• Tag bloat. The average dealer website is loaded with tracking pixels, OEM-mandated scripts, chat widgets, retargeting tags, and third-party overlays. Each fires on every page, degrading performance. Website vendors who charge thousands per month for a managed platform often cannot even place a Google Tag Manager container in the correct location on the page.
• Security exposure. Websites processing customer personal information, credit applications, and trade-in valuations on outdated platforms represent a liability exposure that should concern any general counsel or insurance provider.

You are spending tens of thousands per month on digital advertising and then sending that traffic to websites that Google itself flags as substandard. It is like buying a Super Bowl ad and directing viewers to a restaurant with a health code violation sign on the door.

PART V: THE OEM SUBSIDY DISTORTION — WHEN FREE IS THE MOST EXPENSIVE OPTION

There is a structural problem in automotive retail technology that almost no one discusses openly, because everyone benefits from pretending it does not exist: OEMs subsidize a significant portion of the dealer technology stack, and that subsidy distorts the entire market in ways that suppress innovation and trap dealers in inferior products.

How the Subsidy Works

Manufacturers negotiate enterprise-wide agreements with preferred technology vendors. They then offer these tools to their franchise network at reduced cost or no cost at all, funded through co-op advertising programs, incentive compliance requirements, or direct OEM subsidies. The dealer sees a discounted or free CRM, a subsidized website platform, a bundled digital retailing tool. The OEM gets standardized data pipelines, brand compliance, and control over the digital customer experience.

On the surface, this looks like a benefit. The dealer saves money. The OEM gets consistency. Everybody wins.

Except everybody does not win. What actually happens is a cascading market failure that mirrors one of the most dysfunctional pricing models in American life: healthcare paid by insurance.

The Healthcare Parallel

In American healthcare, the consumer (the patient) rarely pays the direct cost of services. Insurance covers most of the bill. Because the patient is insulated from the true price, they have no incentive to shop on value. Because the provider knows the patient is not price-sensitive, there is no competitive pressure to reduce costs or improve quality. The result is a system where prices are inflated, quality is inconsistent, transparency is nonexistent, and innovation is stifled because the economic feedback loop between consumer and provider is broken.

Now consider the OEM-subsidized dealer technology stack through this lens:

• The dealer (the patient) does not pay the full cost of the tool. The OEM subsidy covers part or all of it. The dealer's price sensitivity is muted.
• The vendor (the provider) knows the buyer is not shopping on value. The vendor's real customer is the OEM, not the dealer. Product development priorities serve the OEM's data and compliance needs, not the dealer's operational needs.
• There is no competitive pressure to improve. A subsidized tool does not need to be the best tool. It needs to be the OEM-approved tool. Vendors compete for OEM contracts, not for dealer satisfaction. A vendor that wins a manufacturer endorsement has a captive distribution channel regardless of product quality.
• Transparency disappears. When the dealer is not paying full price, they do not scrutinize features, performance, or alternatives the way they would with a $15,000 monthly expense coming directly out of operating profit. The subsidy creates the illusion of a good deal while masking the true cost: inferior technology that cannot compete in an open market.
• Innovation is suppressed. Independent vendors who build genuinely superior products cannot compete against "free." A startup with a better CRM, a faster website platform, or a more effective marketing tool faces an impossible sales conversation: "Your product is better, but the OEM gives us this one for free." The best products lose to the cheapest distribution channel, and the entire ecosystem stagnates.

The Independent Procedure Problem

In healthcare, there is a growing movement toward cash-pay and direct primary care models. Patients who pay out of pocket for procedures often find that the cash price is dramatically lower than the insurance-negotiated price. More importantly, they find that providers competing for cash-pay patients deliver better service, more transparency, and faster innovation. When the economic feedback loop is restored, the market works.

The same dynamic exists in automotive technology, but in reverse. Dealers who step outside the OEM-subsidized ecosystem and select tools based on merit, paying full price but demanding full value, consistently report better outcomes. They get CRMs that actually integrate with their marketing stack. They get websites that pass Core Web Vitals. They get AI tools that see across their entire operation rather than through a single vendor's keyhole.

But the OEM subsidy makes that rational choice feel irrational. Why pay $10,000 a month for a superior CRM when the OEM gives you one for $2,000? The answer is the same reason you might pay cash for an MRI instead of going through insurance: because the "cheaper" option carries hidden costs in inferior outcomes, lost time, and missed opportunities that dwarf the sticker price difference.

The Compounding Effect

The subsidy distortion compounds across the entire stack. The subsidized CRM does not integrate well with third-party tools, so you need additional middleware. The subsidized website platform does not perform well, so you spend more on advertising to compensate for poor organic visibility. The subsidized digital retailing tool does not connect to your preferred lender network, so deals take longer. Each subsidized tool creates friction that requires additional spending to work around.

The total cost of ownership for a "free" OEM-subsidized technology stack, including the workarounds, the lost efficiency, the additional vendors needed to compensate for deficiencies, and the opportunity cost of not having tools that actually work, often exceeds what a dealer would spend on a curated, best-of-breed stack purchased on the open market.

The OEM subsidy does not save you money. It costs you money you cannot see. And worse, it costs you adaptability. When AI disruption arrives and you need to pivot quickly, you discover that your "free" technology stack is a cage built by someone else's priorities.

The OEM Mandate Paradox

OEMs also require specific proprietary tracking tags on every dealer website, mandatory for co-op advertising compliance and incentive eligibility. These tags slow down the website, which hurts search rankings and conversion rates, which reduces the effectiveness of advertising spend, which the OEM then questions. The vendor ecosystem has created a performance degradation loop, and everyone is too busy adding tags to notice the loop is destroying the thing it is supposed to measure.

The Overfuel study recommendation is clear: govern tags by template and event, set performance SLAs with vendors tied to Core Web Vitals budgets. The fact that billion-dollar website platforms need to be told to stop loading unnecessary scripts on every page tells you everything about the sophistication level of the subsidized technology ecosystem.

PART VI: THE VENDOR FRAGMENTATION TAX

The average franchise dealership runs between 10 and 15 separate software tools. Some larger operations run 30 or more. Industry data indicates that dealerships average 6.8 software integrations just to complete a single vehicle sale. Each integration represents a separate vendor contract, a separate data silo, and a separate monthly invoice.

The Anatomy of a Typical Dealer Tech Stack

Walk into a franchise dealership and count the software tools:

• DMS: $3,000-$8,000/month
• CRM: $5,000-$15,000/month
• Website platform: $2,000-$5,000/month
• Inventory management/pricing: $1,000-$3,000/month
• Digital retailing: $1,500-$4,000/month
• Chat/AI chatbot: $500-$2,000/month
• Reputation management: $300-$1,000/month
• Service scheduling: $500-$1,500/month
• Equity mining: $1,000-$3,000/month
• Desking: $500-$2,000/month
• F&I menus: $500-$1,500/month
• Call tracking: $300-$1,000/month
• Marketing automation (separate because CRM cannot do it): $500-$2,000/month
• Trade-in tools: $500-$1,500/month

A mid-size dealer is spending $15,000 to $50,000 per month on software before a single integration fee is paid. A ten-rooftop group could easily exceed $200,000 per month on technology, and not one of these systems talks natively to the others.

Every one of these tools exists because the DMS and CRM cannot or will not provide the functionality. The equity mining tool exists because the CRM cannot identify customers in positive equity. The chat tool exists because the native chat is inadequate. The separate email tool exists because the CRM cannot send a campaign. The call tracking tool exists because neither the CRM nor the website can attribute phone calls to marketing campaigns.

You are not running a technology stack. You are running a Rube Goldberg machine held together with duct tape and API calls that fail four out of ten times.

The Bolt-On AI Problem

Every vendor in automotive retail is racing to announce AI capabilities. But every vendor is building AI within its own walled garden. Each vendor's AI sees only its own data. No vendor's AI sees the complete picture of your customer.

A truly intelligent system would need DMS transaction data, CRM interaction history, website browsing behavior, service scheduling data, inventory availability, and marketing attribution. No single vendor controls all these sources. The integration barriers mean no AI system can access all of them in real time.

They are selling you AI that can see through a keyhole when what you need is AI that can see the whole room. And each vendor is absolutely certain that its keyhole is the most important one.

The Human Cost

Your salespeople spend their day toggling between four or five applications. Your BDC agents copy and paste information between systems because integrations do not work. Your service advisors re-enter customer information that already exists elsewhere because tools cannot pull it automatically.

The average salesperson still sells approximately 8-12 cars per month. Digital retailing platforms have been adopted by 91% of dealerships, yet the average customer still spends approximately 3 hours in the store. The technology was supposed to make people more productive. Instead, it created more complexity that required more people to manage. The technology stack is not a solution. It is the problem.

PART VII: THE AI AGENT THREAT — WHAT CHANGES WHEN THE CUSTOMER SENDS A MACHINE

The End of Information Asymmetry

The franchise dealership model has always depended on information asymmetry. You know the invoice price. The customer does not. You know the holdback, the dealer cash, the manufacturer incentives. AI agents eliminate this entirely. An AI agent with market data can determine fair market value, calculate your likely cost basis, identify all incentives, and generate a target price in seconds. Harvard Business Review, Bain & Company, and BCG have all confirmed in early 2026 that AI shopping agents are a present-tense capability, not theoretical.

F&I: The $3,000 Per Vehicle Vulnerability

F&I works because of three factors that AI agents directly undermine:

• Time pressure. The customer wants to leave. An AI agent has infinite patience and comparison-shops before the customer arrives.
• Information complexity. The average consumer cannot compare gap insurance across providers. An AI agent evaluates every product, from every provider, in seconds.
• Cognitive overload. F&I products are presented at maximum cognitive depletion. An AI agent does not experience depletion.

And here is the connection to the technology stack: if your CRM cannot send a targeted email to F&I customers about new offerings, and your website cannot present F&I products transparently to build trust before arrival, you are doubly disadvantaged. You are fighting an AI-armed consumer with tools that cannot perform basic marketing.

Fixed Operations: The Fortress That May Not Hold

Fixed operations is the most defensible profit center. But AI disrupts it by redirecting where work is performed. OEMs are deploying predictive maintenance that flows from the connected vehicle to their cloud, bypassing the dealer. And if your fragmented systems cannot proactively reach customers approaching service intervals, the customer who should have come to you goes to the independent shop because your technology never triggered the reminder.

PART VIII: THE CONSOLIDATION ACCELERANT

Dealership buy-sell transactions surged through 2025. The economics are straightforward: larger groups negotiate better vendor terms, spread technology investment across rooftops, and achieve efficiencies single-point operators cannot match. AI disruption, in either scenario, accelerates consolidation.

The critical question is whether consolidation produces better outcomes or merely concentrates the same problems in fewer hands. If acquiring groups run the same closed platforms, consolidation changes ownership of the problem without solving it. Only if accompanied by technology modernization can consolidation be the mechanism through which the industry adapts.

For the single-point dealer principal, consolidation pressure is real. The response is not to fight it but to ensure that if you sell, you sell from technological strength, and if you stay, you have the infrastructure to compete against groups with scale advantages.

PART IX: THE PLAYBOOK — WHAT MUST BE DONE

The Dealer Principal's Survival Checklist

• Audit every vendor contract immediately. Know your termination provisions, data portability rights, integration costs, and renewal timelines. Use approaching renewals as leverage.
• Evaluate cloud-native DMS alternatives. The switching cost is significant, but the cost of not switching may be higher. New entrants have delivered thousands of features in a fraction of the time incumbents take.
• Demand CRM interoperability or leave. If your CRM cannot integrate with standard marketing automation tools, advertising platforms, and third-party data sources via open APIs, it is not a CRM. It is a data jail.
• Question the OEM subsidy. Calculate the total cost of ownership for your subsidized tools, including workarounds, lost efficiency, and additional vendors needed to compensate. Compare that to what a curated, best-of-breed stack would actually cost. The "free" option may be the most expensive choice you are making.
• Take control of your website. Demand to know what software versions your website platform runs. If the answer involves versions that are years behind current releases, you are competing with outdated tools. Evaluate modern alternatives and set performance SLAs tied to Core Web Vitals.
• Audit your tag situation. Run PageSpeed Insights on your site. If your Core Web Vitals are failing (and the Overfuel study says there is a 99.6% chance they are), demand fixes and performance accountability.
• Consolidate your vendor stack ruthlessly. Every tool that exists because another tool cannot do its job is a symptom of a broken ecosystem. Map your stack, identify redundancies, and eliminate.
• Invest in F&I transformation now. Move toward transparent pricing and advisory-based selling before AI agents force it.
• Double down on fixed operations. Service is the most defensible profit center. Invest in technician training, mobile service, and customer experience.
• Build a data strategy. Know what data you generate, where it resides, who accesses it, and what it costs. If your DMS vendor charges you $42,000 a year to access your own data, plan to eliminate that cost.
• Hire AI-literate leadership. The general manager of 2028 needs to understand AI strategy and data architecture the way the GM of 2018 needed to understand digital marketing.

For Vendors and OEMs

The vendor community faces a choice: continue as gatekeepers, extracting integration revenue from closed ecosystems, or transform into platform providers enabling open interoperability. The gatekeeper model has been profitable, but it is built on friction. The market will answer this question faster than most vendors expect.

Manufacturers should mandate open data standards for the franchise network, invest in dealer technology modernization rather than just customer-facing digital retail, build AI strategies that include the dealer rather than routing around them, rationalize their tag requirements, and critically examine whether their technology subsidy programs are helping or hindering the dealers they are supposed to support. Connected vehicle data should flow to the dealer, not around the dealer.

PART X: THE TIMELINE — WHAT HAPPENS WHEN

CONCLUSION: THE CANARY IS STILL ALIVE, BUT THE MINE IS GETTING DARKER

Citrini ends their scenario with a haunting observation: the canary is still alive. We are in early 2026, not in the fictional crisis of June 2028. There is time. Citadel argues the data does not support panic but acknowledges AI will alter work and demand.

For automotive retail, the synthesis yields a clear imperative: prepare for the Citrini scenario while operating under the Citadel timeline. Invest now in open technology infrastructure, transparent business models, strengthened service operations, AI-literate leadership, and data strategies that prioritize your control.

But here is the part that neither Citrini nor Citadel could know, because they are not dealer principals: the biggest threat to your dealership is not AI. It is the technology infrastructure that is supposed to help you respond to AI. Your DMS charges you to access your own data. Your CRM cannot perform basic marketing functions. Your website platform may be running years behind on security and performance. Your vendor stack includes a dozen or more tools that exist only because the core systems cannot do their jobs. And underneath it all, OEM subsidies distort the market, suppressing the innovation that could save you, exactly the way insurance-mediated pricing suppresses innovation in healthcare.

The franchise dealer system has survived for over a century by adapting. It survived the internet, Tesla, and COVID. But AI is different. AI gives consumers agency. An AI agent negotiates, comparison-shops, optimizes. It does not forget, does not get tired, and does not feel social pressure. Every friction-based revenue stream is exposed to an adversary specifically designed to eliminate friction.

The franchise dealer system can survive this. But it cannot survive it running on technology infrastructure designed to monetize friction rather than eliminate it.

Audit your contracts. Demand open APIs. Question the OEM subsidy. Evaluate modern alternatives. Consolidate your vendor stack. Invest in the technology infrastructure that will let you compete in an AI-driven market.

The intelligence crossroads is here. The road you take from this point forward will determine whether your dealership in 2030 is a thriving, technology-enabled business or a cautionary tale.

Choose wisely. And choose quickly. The AI agents are not waiting for your vendor to finish bolting on features.

Credit

Published by Dealer AI Guy. This document was written with research and editorial assistance from Claude, Anthropic’s AI assistant.

Sources and Methodology

This analysis draws on the following primary and secondary sources: Citrini Research, "2028 Global Intelligence Crisis" (February 2026); Citadel Securities, "2026 Global Intelligence Crisis" (February 2026); CDK Global antitrust settlement filings ($129.5M combined, February 2025); Overfuel Core Web Vitals study of 1,910 dealer websites (September 2025); Presidio-NCM Average Dealership Performance Benchmark Q4 2025; WordPress.org version history and release notes; Ford Motor Company Signature 2.0 retail experience announcement (November 2025); STAR Retail Automotive Domain Model (January 2026); DealerRefresh industry forums and surveys; vendor pricing from publicly available documentation; and direct dealer experience.