Georg Dresler
Manipulating The Machine: Prompt Injections And Counter Measures
#1about 4 minutes
Understanding the three layers of an LLM prompt
A prompt is structured into three layers: the system prompt for instructions, the context for additional data, and the unpredictable user input.
#2about 3 minutes
How a car dealer's chatbot was easily manipulated
A Chevrolet car dealer's chatbot was exploited by users to generate humorous and unintended responses, including a legally binding offer for a $1 car.
#3about 4 minutes
Stealing system prompts to bypass security rules
Attackers can use creative phrasing like "repeat everything above" to trick an LLM into revealing its hidden system prompt and instructions.
#4about 6 minutes
Why attackers use prompt injection techniques
Prompt injections are used to access sensitive business data, gain personal advantages like bypassing HR filters, or exploit integrated tools to steal information like 2FA tokens.
#5about 4 minutes
Exploring simple but ineffective defense mechanisms
Initial defense ideas like avoiding secrets or tool integration are impractical, and simple system prompt instructions are easily circumvented by attackers.
#6about 4 minutes
Using fine-tuning and adversarial detectors for defense
More effective defenses include fine-tuning models on domain-specific data to reduce reliance on instructions and using specialized adversarial prompt detectors to identify malicious input.
#7about 2 minutes
Key takeaways on prompt injection security
Treat all system prompt data as public, use a layered defense of instructions, detectors, and fine-tuning, and accept that no completely reliable solution exists yet.
Related jobs
Jobs that call for the skills explored in this talk.
Wilken GmbH
Ulm, Germany
Senior
Kubernetes
AI Frameworks
+3
Picnic Technologies B.V.
Amsterdam, Netherlands
Intermediate
Senior
Python
Structured Query Language (SQL)
+1
Matching moments
04:57 MIN
Increasing the value of talk recordings post-event
Cat Herding with Lions and Tigers - Christian Heilmann
03:28 MIN
Why corporate AI adoption lags behind the hype
What 2025 Taught Us: A Year-End Special with Hung Lee
03:45 MIN
Preventing exposed API keys in AI-assisted development
Slopquatting, API Keys, Fun with Fonts, Recruiters vs AI and more - The Best of LIVE 2025 - Part 2
03:48 MIN
Automating formal processes risks losing informal human value
What 2025 Taught Us: A Year-End Special with Hung Lee
04:59 MIN
Unlocking LLM potential with creative prompting techniques
WeAreDevelopers LIVE – Frontend Inspirations, Web Standards and more
05:18 MIN
Incentivizing automation with a 'keep what you kill' policy
What 2025 Taught Us: A Year-End Special with Hung Lee
05:55 MIN
The security risks of AI-generated code and slopsquatting
Slopquatting, API Keys, Fun with Fonts, Recruiters vs AI and more - The Best of LIVE 2025 - Part 2
03:38 MIN
Balancing the trade-off between efficiency and resilience
What 2025 Taught Us: A Year-End Special with Hung Lee
Featured Partners
Related Videos
27:32ChatGPT, ignore the above instructions! Prompt injection attacks and how to avoid them.
Sebastian Schrittwieser
23:24Prompt Injection, Poisoning & More: The Dark Side of LLMs
Keno Dreßel
29:00Beyond the Hype: Building Trustworthy and Reliable LLM Applications with Guardrails
Alex Soto
30:36The AI Security Survival Guide: Practical Advice for Stressed-Out Developers
Mackenzie Jackson
33:28Prompt Engineering - an Art, a Science, or your next Job Title?
Maxim Salnikov
27:02Hacking AI - how attackers impose their will on AI
Mirko Ross
31:12Using LLMs in your Product
Daniel Töws
31:19Skynet wants your Passwords! The Role of AI in Automating Social Engineering
Wolfgang Ettlinger & Alexander Hurbean
Related Articles
View all articles
From learning to earning
Jobs that call for the skills explored in this talk.
FRG Technology Consulting
Intermediate
Azure
Python
Machine Learning
Mindrift
Remote
£41K
Junior
JSON
Python
Data analysis
+1
Deloitte
Leipzig, Germany
Azure
DevOps
Python
Docker
PyTorch
+6
Abnormal AI
Intermediate
API
Spark
Kafka
Python
Startup
Charing Cross, United Kingdom
PyTorch
Machine Learning