Data retention and mass surveillance
The discussion opened with a fundamental question: can traditional, targeted investigative measures still function in a digital environment where crime is borderless, fast, and often anonymous? The moderator framed the issue by contrasting historical approaches such as wiretapping specific suspects with contemporary methods for population-wide data retention and scanning.
Responding to this shift, Olivier Leroux argued that the digital age has altered long-standing assumptions about investigative proportionality. Decades ago, large national DNA or fingerprint databases were seen as incompatible with the presumption of innocence. Today, he suggested, the nature of digital crime requires us to reassess. In his view, mass data retention is no longer a convenience for prosecutors but a necessity, as data is often the only trace left behind or even the crime itself. He acknowledged that the Court of Justice of the European Union has raised concerns and imposed strict criteria on data storage but noted that aligning these legal requirements with investigative reality is extremely difficult. Expecting both perfect privacy and perfect security, he argued, is unrealistic.
Irene Kamara took a more cautious position. While acknowledging the operational challenges investigators face, she stressed that privacy and confidentiality of communications are foundational to democracy and the rule of law. “These protections are not optional”, she said, “but essential safeguards, particularly for journalists, whistleblowers, and individuals living under less democratic regimes.” Kamara also questioned the assumption that more data automatically leads to better investigations. Large datasets, she warned, often generate noise rather than insight, overwhelming already stretched law enforcement capacities. From her perspective, the real need is not more data, but better-quality, relevant, and reliable data.
Both speakers converged on the point of fragmentation. The current patchwork of national rules across Europe creates legal uncertainty and uneven safeguards. A harmonised European framework, they agreed, could help balance investigative needs with fundamental rights more consistently.
End-to-end encryption and the ‘back door’ debate
The second topic tackled one of the more polarising issues: encryption. End-to-end encryption protects journalists, activists, and ordinary citizens, yet law enforcement agencies increasingly warn they are “going dark” as crucial evidence becomes inaccessible. Moderator Charlotte Conings asked whether technology providers should be obliged to build “back doors” into encrypted services, allowing authorities access under exceptional circumstances.
Irene Kamara responded unequivocally that this would be a mistake. “Once a back door is opened”, she argued, “there is no way back”. Opening such access would permanently weaken the confidentiality of communications for everyone, including politicians and journalists. Instead of breaking encryption, she suggested alternative approaches: “In countries like the UK and the Netherlands targeted preventive measures are taken, with the police directly warning individuals downloading illegal material, demonstrating that anonymity is an illusion.” Kamara also pointed to regulatory tools such as the Digital Services Act, which can compel large platforms to assess and mitigate how their services are misused.
Olivier Leroux acknowledged the cybersecurity risks but expressed that concerns often overlook the growing power of Big Tech companies effectively controlling encryption and data. He argued that there is a risk of creating a parallel digital space beyond oversight from the law. He cited the Sky ECC case as an example of why investigative access matters: by breaking encrypted communications, authorities were able to “sit at the table” with organised criminals who believed they were invisible. In Belgium, he noted, “legislation already allows suspects to be charged for refusing to provide access to encrypted data. “Offering criminals a guaranteed safe space in the digital world would be a mistake,” he argued.
Despite their differences, both speakers recognised the need for careful limits. The debate revealed a shared concern that any solution must avoid undermining cybersecurity, while still ensuring that lawful investigations are still possible.
The rise of AI-driven police work
The final section examined the growing role of artificial intelligence in criminal justice. While criminals already exploit AI tools, law enforcement is only beginning to integrate such new technologies, raising questions about bias, transparency, and accountability.
Moderator Charlotte Conings asked how far authorities should go with predictive algorithms and automated analysis. Olivier Leroux observed that AI is not yet widely used in criminal investigations but predicted that “facial recognition would soon become an important factor.” He pointed to rapid technological developments in countries like China, noting that while such systems may offer investigative benefits, they raise profound privacy concerns.
Irene Kamara expanded on these risks by highlighting how criminals already use AI to manipulate images, remove backgrounds, and conceal identities. From her perspective, law enforcement relying on AI is inevitable, but it carries the danger of accepting algorithmic outputs uncritically. The problem, she explained, is compounded by the fact that many AI systems are privately owned and commercially exploited, making independent audits difficult or even impossible.
To mitigate these risks, Irene Kamara advocated for developing “homegrown” tools within trusted authorities, trained on high-quality police data. “Rather than importing systems from private vendors”, she argued, “law enforcement should prioritise transparency, accountability, and human oversight.”
The moderator concluded the debate by noting a shared concern: technologies introduced to fight serious crime or terrorism often expand in scope over time. Without strong safeguards, tools designed for exceptional cases may gradually become routine. The discussion, she suggested, underlined the importance of setting up clear legal and ethical boundaries before such technologies are widely deployed.
Across all three topics, the debate revealed a consistent tension between effectiveness and restraint. Whether discussing mass data retention, encrypted communications, or AI-driven policing, the panellists agreed that adapting to digital crime is unavoidable. Yet they also stressed that security measures must remain grounded in democratic principles, legal certainty, and proportionality. The challenge ahead lies not in choosing between privacy and security, but in designing frameworks that credibly protect both.
Picture 1: Irene Kamara
Picture 2: Charlotte Conings
