Artificial intelligence is profoundly transforming corporate business models. It does more than simply enhance the efficiency of existing processes. It is reshaping how value is created, captured, and monetized. Here are a few key trends that stand out:

1. The Disintermediation of Revenue Channels

AI enables companies to bypass many intermediaries that have traditionally captured a significant share of value. In finance, particularly in consumer lending, platforms such as Upstart use AI models to assess customers’ creditworthiness based on richer data sets (education history, payment behavior, online activity, etc.) that go beyond traditional FICO credit scores. The results: more approved applications (up 27% according to the Consumer Financial Protection Bureau) at lower rates (16% lower APR on average), with comparable risk levels. Where banks historically controlled the entire chain, from credit scoring to pricing, they are now reduced to liquidity providers, while the AI-native platform captures the customer relationship, the data, and the associated margins.

AI is also giving brands the ability to sell directly. Previously, an SME wishing to market its products had to rely on distributors or third-party marketplaces that took a substantial share of the margin. Today, with tools like Shopify Sidekick or Amazon’s generative AI features, a brand can automatically create product listings, email campaigns, visuals, and manage its sales strategy without needing an agency or marketing team.

In digital advertising, where the value chain was traditionally fragmented among creative agencies, media agencies, ad networks, and tracking tools, major platforms have verticalized the process through AI. Tools like Google Performance Max, Meta Advantage+, TikTok Smart Campaigns, and Amazon Ads allow SMEs to create complete campaigns in just a few clicks: covering targeting, creation, distribution, and optimization. The AI generates visuals, selects channels, and adjusts budgets in real time.

2. The Creation of New Products and Services

AI is enabling the emergence of offerings that simply did not exist before. The automotive industry provides a striking example. Historically, carmakers’ business models relied on a one-time vehicle sale, capturing most of the margin at delivery, with only limited additional revenue from financing or maintenance. Once the car was sold, the commercial relationship largely ended.

Today, with the growing integration of embedded AI, companies like BMW and Mercedes-Benz are transforming the car into a software platform. Certain features (intelligent climate control, driver assistance, or engine performance optimization) are now offered as monthly subscriptions or post-purchase upgrades. The vehicle thus becomes an entry point into a freemium model, where the hardware serves as a foundation for ongoing software monetization.This shift profoundly transforms manufacturers’ revenue structures. It creates recurring income streams, extends customer lifetime value, and allows for finer segmentation of offerings.

3. Dynamic and Personalised Pricing

Before AI, pricing in many industries (transportation, e-commerce, hospitality, and insurance etc.) followed relatively rigid frameworks based on general rules: periods of high demand, seasonality, or customer segments defined by fixed criteria (age, location, status). Pricing strategies were often adjusted manually at regular intervals or according to basic “if–then” rules.

AI has fundamentally transformed this logic by enabling real-time, personalized, and self-learning pricing models. These systems analyze thousands of signals for each individual (browsing history, purchase intent, estimated income level, price sensitivity, sales channel, and even contextual factors like time of day, weather, or device used).

Tangible examples of AI-native pricing are already operating across several industries:

• Air Transport: Airlines have long relied on sophisticated pricing engines, but AI takes this a step further by incorporating behavioral variables drawn from users’ on-site activity (time spent browsing, multiple destination searches, abandoned carts etc) to adjust displayed prices according to each user’s likelihood of purchase. Solutions developed by companies such as PROS, Amadeus, and Sabre are already being used by carriers such as Lufthansa and Emirates.
• E-commerce: Platforms such as Amazon and Zalando use AI to continuously adjust product prices based on demand fluctuations, browsing behavior, inventory levels, and customer profiles.
• Insurance: In the past, insurance premiums were calculated once a year based on average statistical data. Some products now leverage policyholder-specific behavioral data to adjust risk premiums dynamically. Thanks to AI, embedded sensors in vehicles can now measure actual driving behavior (hard braking, speeding) to continuously assess real risk and reward safe driving. Pricing thus no longer depends solely on the probability of an incident, but also on the user’s active efforts to prevent it.

AI enables capturing more value at the individual level by tailoring prices not to average segments, but to each customer within their specific context. This dynamic also opens up new possibilities for variable or adaptive pricing models that can evolve over time (smart subscriptions or usage-based billing) and strengthen the alignment between perceived value and price paid.

What will be the impact of AI on Corporate Social Responsibility? While it represents a tremendous productivity driver, it also relies on energy-intensive infrastructures and the massive use of data, sometimes at the expense of copyright laws.

Like any systemic technology, AI can be a powerful driver of progress, but it also creates new risks that must be acknowledged and carefully managed.

There are already many positive examples. In the hospitality sector, Accor uses AI to adjust heating and air conditioning in its establishments based on occupancy rates, reducing energy consumption by up to 20%. In the energy sector, RTE (Réseau de Transport d’Électricité – Electricity Transmission System Operator) employs AI to anticipate consumption peaks and balance supply and demand in real time, including the integration of renewable energy sources. In logistics, Amazon optimizes its routes and truck load capacities to reduce unnecessary trips and limit emissions related to last-mile delivery, achieving measurable gains in carbon footprint per package delivered. AI also promotes accessibility; automatic translation and subtitling open up information to audiences previously excluded, such as those who are hearing or visually impaired. Finally, in healthcare, startups like Qure.ai in India deploy AI-powered medical image analysis tools to automatically detect cases of tuberculosis or pneumonia, even in the absence of radiologists, thus improving diagnostic access in underserved areas.

These advances should not overshadow the fact that AI, especially in its generative form, relies on highly energy-intensive infrastructures. Training a single large language model can require several hundred tons of CO₂ equivalent, depending on the data center and the source of electricity used. A simple request via a chatbot can consume up to nine times more energy than a standard Google search.

On the ethical front, AI raises fundamental questions. Taking insurance as an example, how far can we go in monetizing compliance with “good behavior”? At what point does this amount to a disguised form of discrimination?

Debates around responsibility and consent are highly complex. The issues surrounding copyright and the use of journalists’ work without consent for AI training raise serious challenges regarding intellectual property. Can a business model be built on content taken without compensating the creators? More broadly, data collection and usage remain largely opaque and often occur without informed consent, with users not always aware of how their data might affect their access to a service or the prices they are offered. Even more concerning, several recent cases have highlighted the risks associated with prolonged interactions with chatbots, underscoring the importance of safeguards to ensure psychological safety.

One of the major challenges posed by generative AI is that of responsibility. At its core, the central question is governance. Who should have access to what? Who should be compensated, and who is accountable for decisions made by an algorithm? The company? The model developer? The cloud provider? It is not the capabilities of AI that are problematic, but the collective framework we choose, or fail, to establish around it.

To move forward, several levers are essential. These include strengthening data traceability and model explainability, ensuring human oversight in sensitive sectors, conducting regular audits to identify and address biases, and above all, securing clear user consent regarding the use of their data. AI forces us to confront a fundamental question: what kind of society do we want to live in?

Beyond the immediate transformations, what structural changes do you anticipate for business models as a result of AI?

AI could, in many respects, follow an impact trajectory similar to that of the Internet, but with even greater speed and depth. In the 1990s, the Internet first transformed communication and information channels (email, websites), then disrupted distribution chains (e-commerce), relationship models (social networks), and eventually imposed an entirely digitized economic logic. We can expect AI, beyond its initial visible effects (automation of support functions, productivity gains, enhanced customer experience), to drive similar transformations, this time at the level of judgment, decision-making, and even the architecture of business models.

In the near future, generative AI will be integrated into almost every business environment (writing, summarizing, reporting, forecasting) with contextual assistants capable of producing, recommending, and adjusting in real time, closely aligned with daily decision-making. But as models become more powerful, this integration will extend beyond “support functions”. AI will take a central role in strategic decision-making processes (portfolio analysis, resource allocation, pricing, scenario modeling), with systems capable of independently generating options or high-level recommendations.

This will likely trigger a twofold structural shift. On one hand, a concentration around a few major AI infrastructure providers (foundational models, cloud platforms, frameworks) who will play a systemic role, much like the GAFAM did after the rise of the Internet. On the other hand, there will be sector-specific fragmentation, with the emergence of specialized AI solutions tailored by industry, usage, or even geography, deeply integrated into business processes.

As AI consumes ever more resources (data, computing power, and energy) we can also expect the rise of new performance metrics that incorporate not only profitability but also the environmental footprint and the algorithmic sustainability of a business model.

In the longer term, AI could also redefine the concepts of work and governance. Just as the Internet dematerialized exchanges, AI may dematerialize certain forms of decision-making, or even management itself. The question will no longer be, “Which tasks can be automated?” but rather, “In which areas of collective responsibility should AI remain a tool, and not become the decision-maker?”

Following the digital transformation sparked by the Internet, AI ushers in a cognitive transformation. It challenges who thinks, who decides, and on what grounds. This is not just a technological revolution; it is a revolution in governance.

The Internet, in its time, enabled major advances but also became a channel for illegal, dangerous, and unethical uses. And despite more than twenty years of hindsight, legal frameworks and regulatory mechanisms remain incomplete. The parallel is enlightening. It is not about “stopping” AI any more than it would have been realistic to stop the Internet. Rather, it is about defining now the safeguards, principles of use, and goals we consider desirable.

The issue is not merely technical or economic. AI forces us to make collective choices about what we want to preserve, encourage, or prohibit and to fully acknowledge that behind every algorithm there is an intention, or a governance gap.