Key Messages of the 2025 Octopus Conference related to AI, Cybercrime and Criminal Justice
The 2025 Octopus Conference on Cooperation against Cybercrime was held in Strasbourg from 4-6 June 2025. There were some relevant and interesting sessions on cybercrime, e-evidence and AI, pig butchering scams, crypto investigations, cyberviolence and child sexual abuse, cybercrime as a war, et.al. some of which were mainstreamed and will soon be available.
During the last day of the conference, the Council of Europe released the key messages which provides a summary of each of the sessions that took place during the three days and the main takeaways of the discussions held in each session. Concerning the key messages intersecting with Artificial Intelligence and Criminal Justice that came out of this conference are two:
“Artificial intelligence (AI) is reshaping cybercrime, both in terms of offences (against, by and by means of AI systems) and in terms of opportunities to investigate crime and collect electronic evidence. This raises a number of complex legal and practical questions, including the question of the applicability of treaties such as the Convention on Cybercrime and its Protocols. The Octopus session on AI underscored that the criminalisation provisions and its cooperation tools under the Budapest Convention framework are likely to be applicable to a varied typology of AI-related scenarios. In this respect, the mapping study currently being prepared by the Working Group on Artificial Intelligence of the Cybercrime Convention Committee (T-CY) is expected to offer further insights and a more in-depth analysis of these issues. Capacity building to permit criminal justice authorities to address challenges related to AI is needed”.
“AI-generated child sexual abuse material (CSAM) remains a serious and growing threat to children requiring an effective criminal justice response. Legislative and policy gaps persist in some jurisdictions. Prosecution is often hindered by definitional ambiguities, evidentiary challenges, and the cross-border nature of online abuse. The harm caused by AI-generated CSAM – whether to survivors of past abuse or to societal perceptions of child sexual exploitation and sexual abuse – is real and significant. Addressing this threat requires legal definitions to be adapted to include synthetic content, investigators to be equipped with tools to distinguish AI-generated from CSAM featuring an identifiable child victim, and international cooperation to be strengthened through instruments such as the Lanzarote Convention and the Budapest Convention on Cybercrime, including its Second Additional Protocol. Leveraging AI for detection and analysis, while ensuring strong safeguards and oversight, is also critical to an effective and child-centred response”
It is worth mentioning that I followed some of the session including Main session 4 on cybercrime, evidence and AI, where there were some interesting interventions particularly from the representative of Microsoft who talked about the role of this company to identify ongoing threats, the challenges to identify attacks against AI providers and the complexities and importance of the cooperation with law enforcement authorities to tackle cyber enable crimes and AI assisted crimes.
Further, the representative of the Committee on Artificial Intelligence (CAI) discussed the importance of the protection of fundamental rights in the context of the development of AI systems, the HUDERIA methodology developed by the CAI and the challenges that lie ahead for the countries that sign and ratify the Council of Europe Framework Convention on Artificial Intelligence and human rights, democracy and the rule of law including the monitoring of the application of said convention.
Finally, it was mentioned about the ongoing work of the Working Group on Artificial Intelligence created within the Cybercrime Convention Committee (T-CY) which will prepare a mapping study this year that would seek to explain and clarify how the Budapest Cybercrime Convention and its two additional protocols could apply to AI systems, including definitions, criminal conducts and typologies and procedural measures including the future work of the State Parties to the Convention on Cybercrime (currently 80 States) in this area, which I look forward to reading and particularly to see how it would be implemented in practice among the State signatories of the Budapest Cybercrime Convention considering that this area is moving extremely fast and requires a more dynamic and consistent approach specially in facilitating assistance and training to strengthen the capacities of criminal justice authorities responsible for investigations and for the implementation of AI systems that could alsohelp to facilitate, enhance and improve their work.
#AI, #BudapestConvention #cybercrime #ELPACCTO2.0 #Europol #Eevidence #electronicevidence #EUROJUST #INTERPOL #Internationalcooperation #justice #Octopus #Organizedcrime #phishing #pigbutchering #Ransomware #RasS #RuleofLaw #Safeguards #SecondAdditionalProtocol #SIRIUS #StrongerTogether #UNODC