Generative AI (GenAI) is transforming how competition-related cases, including merger control and investigations, are handled. Its ability to analyse vast amounts of data quickly and accurately is proving to be a game-changer, particularly in high-stakes scenarios like cartel investigations, where speed and precision are critical.

Companies across sectors and jurisdictions are increasingly turning to AI to ensure they meet their competition regulatory challenges. This is particularly relevant in times when antitrust authorities in the U.K. and Europe are expanding the breadth and scale of their investigations, and increased expectations for companies to analyse substantial data and deliver accurate responses promptly.

It is not just companies that are embracing AI. Competition authorities in Europe are also warming up to its speed-and-scale potential. Spain’s antitrust watchdog has been particularly active in deploying AI solutions to detect collusive bids in public procurement processes.¹ A recent General Court judgment also confirmed the European Commission’s use of AI-assisted tools in analysing patterns in public communications, from earnings calls to strategic statements, to detect evidence of collusion.² The Court clarified that “sufficiently serious indicia” and clear articulation of suspicions are required, while implicitly validating AI as a tool to detect possible infringements.

Increasing recognition from authorities of the technology’s benefits signals their growing acceptance of the use of AI tools across the industry in competition cases, where time is of the essence.

Efficiency and speed in competition cases

When allegations arise in competition cases, especially cartel investigations, companies must race to understand their legal position and respond accurately to regulators, often in extremely tight timeframes.

GenAI’s capacity to process hundreds of thousands of documents from various sources within 24 to 48 hours provides a clear picture of the situation a lot faster than traditional methods. At A&M, we have successfully deployed GenAI tools for reviewing large volumes of documents in dozens of cases, with remarkable results in terms of speed and efficiency. This speed is crucial for making informed decisions and developing the best legal strategies to cooperate with authorities while presenting a robust defence against allegations.

For example, A&M was recently approached for our services in a case where three companies were raided, facing allegations of cartel activity. By applying GenAI to available data on a Friday night, we were able to analyse the information and understand our client’s position before the weekend was up. Previously, this process would have taken weeks, potentially losing our client the legal advantage.

In competition cases, being the first to understand your position spells the difference between securing leniency or facing significant penalties. Leniency programs allow only one company to "blow the whistle" and receive reduced fines or immunity, making speed a critical factor.

Applications of GenAI in competition law

GenAI is being used in several key areas of competition cases to drive efficiency and improve outcomes:

1. Document review and analysis

GenAI excels at reviewing large volumes of documents, identifying relevant materials and flagging potential issues speedily. For example, it can quickly identify key documents, helping you to understand your position faster than ever. It can also help in finding privileged material, ensuring that sensitive or confidential information is not inadvertently handed over to regulators. This is particularly important in merger control cases, where companies are often required to submit vast amounts of data under tight deadlines. AI tools can strip out privileged material efficiently, providing greater confidence in the data being shared.

2. Case strategy and reporting

GenAI can assist in building case strategies by organising key documents, creating chronologies, and even drafting initial reports. For instance, it can analyse documents related to a specific individual and generate a first draft of a report explaining their involvement in a case. This functionality streamlines the preparation process and allows legal teams to focus on higher value tasks.

3. Short-form communication analysis

Traditional algorithms often struggle with short-form communications and messages, due to the lack of extensive text. GenAI is far more effective at analysing these types of communications, including emojis in context, making it a valuable tool in investigations where such messages are critical evidence.

4. Proactive compliance

While not yet widely adopted, companies can use GenAI proactively for compliance purposes. For example, AI can scan internal communications for anti-competitive language or conduct periodic compliance reviews by sampling data from key custodians. This approach, which was previously time-consuming and costly, is now more feasible due to the efficiency and cost-effectiveness of AI tools.

Prompt engineering: An essential skill

A tool is only as good as the user’s mastery of it. A key success factor in using GenAI is prompt engineering—the process of crafting precise and effective prompts to guide the AI system. The quality of the output depends heavily on the quality of the input.

Here at A&M, we have developed a comprehensive library of prompts tailored to different case types and are poised to adopt any new features rolled out by AI solution providers. With the right prompts, we have consistently reduced review timeframes, often from weeks to just days, and seen substantial cost savings.

Our approach, a blend of AI and human expertise, consists of three elements: First, AI specialists set up cases, organise case materials and develop sample prompts, iterating the process and working with the client’s legal team and subject matter experts until we have a very good sample set. Then, the AI platform generates results and lastly, we quality check the sample outputs before we are ready to apply to the full document set.

A well-drafted prompt can instruct the AI to review documents, identify privileged material, or generate a report, delivering highly accurate and relevant results. This expertise in interacting with the technology is what sets experienced users apart, much like an expert in Microsoft Excel can achieve far more in Excel than a casual user.

Dawn raids and the role of AI

GenAI is also proving invaluable in a key aspect of competition enforcement - dawn raids, where regulators raid company premises to collect data for investigations. The exact figures for dawn raids are difficult to track across Europe, with different regulators following different publication rules, but we know of at least 13 raids across Europe in the first quarter of 2025.³ The actual number is likely to be higher. When a raid occurs, a company’s response is likely to include their own investigation into the alleged issue, with the need to quickly analyse the seized data to understand their position. GenAI enables this rapid analysis, providing a strategic advantage in responding to regulators.

In a recent dawn raid case, we were able to deploy AI tools to analyse the data within hours, giving our client the option to either cooperate with regulators or defend their position. Without this speed, the client could have lost the opportunity to secure leniency or develop an effective defence strategy.

Regulatory Acceptance of Generative AI

One common question is whether competition regulators allow the use of Gen AI by companies in their response to the regulator’s investigation. While there is no formal guidance or judgments yet from the competition authorities on how companies can deploy these technologies to analyse data in competition cases, we have enough evidence to show that, in the right hands, GenAI technologies are producing more accurate results than the traditional human review process. Acceptance is growing and will be formally reported on soon. In these matters, clearly defining the approach taken, and verifying the results produced by GenAI using traditional methods such as statistical sampling and subject matter expert sign off, is critical.

Conclusion

GenAI is revolutionizing competition cases by driving efficiency, reducing risks and speeding up analysis, often by several orders of magnitude, enabling more thorough and informed decision-making in time-sensitive scenarios. From document review and privilege identification to case strategy and compliance, its applications are vast and impactful. The key to unlocking its full potential lies in expertise—particularly in prompt engineering and understanding how to interact with the technology and when to step in with human oversight. With regulators also starting to adopt the technology for scrutiny, companies must leverage GenAI effectively to strategically navigate the complexities of competition law, or risk getting left behind.