Hacker Summer Camp in Las Vegas hosts several conferences focused on diverse aspects of information security, including application security, cloud security, red teaming, and AI security. Datadog researchers participate actively by delivering talks and open source tools while providing resources for remote attendees. #HackerSummerCamp #Datadog #AIsecurity
Keypoints
- Hacker Summer Camp consists of multiple conferences: The Diana Initiative, BSides Las Vegas, Black Hat USA, and DEF CON, each targeting different audiences and topics.
- Conference themes span application security, cloud security, red teaming, applied research, supply-chain security, and threat intelligence.
- AI security is a major focus this year, covering AI model attacks, AI-assisted threat actor techniques, AI defense mechanisms, and AI ethics.
- Datadog researchers will be actively participating, presenting talks and sharing open source tooling with the community.
- A curated list of the top 50 talks at Hacker Summer Camp 2025 is available with categorization by theme and includes links to conference sessions.
- Resources such as slides and recordings from the talks will be updated as they become available to support both in-person and remote attendees.
- Community engagement is encouraged through newsletters and RSS feeds to keep updated on the latest research and events.
MITRE Techniques
- [T1499] Endpoint Denial of Service – Related to attacking AI-powered applications potentially causing disruption (“Methods of attacking and abusing AI models and AI-powered applications”).
- [T1204] User Execution – Exploiting prompt injection vulnerabilities in AI systems (“…exploiting prompt injection vulnerabilities or backdooring models”).
- [T1486] Data Encrypted for Impact – Backdooring AI models to alter their integrity (“…backdooring models”).
- [T1598] Phishing – Use of AI to generate phishing campaigns enhancing efficiency (“Threat actors that use AI to be more efficient in their operations, such as generating phishing campaigns”).
- [T1027] Obfuscated Files or Information – Using AI to bypass security mechanisms (“…delivering malicious payloads, or bypassing security mechanisms”).
- [T1616] Application Layer Protocol – Security analysis of AI-generated code by analyzing the output of code-generating large language models (“…analyzing the output of a code-generating LLM for security issues”).
- [T1592] Gather Victim Host Information – Use of AI for threat intelligence and streamlined security operations (“Use of AI for defense, such as being more efficient at finding evil or streamlining security operations”).
Indicators of Compromise
- [File Names] Examples of tooling and talk resources shared by Datadog related to AI security topics – specific file names not listed but reference to open source tooling provided.
- [Domains] Conference websites for The Diana Initiative, BSides Las Vegas, Black Hat USA, and DEF CON provide access to talk resources and recordings – examples include blackhat.com and defcon.org.
Read more: https://securitylabs.datadoghq.com/articles/hacker-summer-camp-guide-2025/