Nick Evans
COO
Lockular

SESSION OVERVIEW: Data will drive the new economy, but the digital world will prosper to the benefit of us all only if people trust data- deep fake and AI make that a real concern.
I’ll explore the problems and give examples, such as evidence presented to UK Parliament recently to demonstrate vulnerabilities of data used in critical infrastructure, and how we all generate a huge volume of data that, unless it’s secured and accounted for, can be used by a bad actor to track our lives second by second.
The solution is offered by innovative web3 technology by writing an immutable record in blockchain, guaranteeing data is complete, accurate and real, thereby not only securing the data, but crucially ensuring its use is accountable and transparent.
Through such solutions, critical infrastructure is secured, we can be sure the data we use in our personal and business lives is complete, accurate and real, and we take control of personal data by ensuring it’s used only for purposes we’ve given consent

Ricardo Sueiras Sueiras
Principal Advocate
Open Source

SESSION OVERVIEW: Authorization is one of the foundational needs when building your applications and services. Making sure you are able to correctly and reliably grant or deny access to your application resources is critical. Yet this is still one of the hardest things we have to do as developers.

In this session I am going to cover a new open source project called Cedar that helps you solve this problem and build next generation authorization into your applications. I am going to provide a quick overview in authorisation first before looking at Cedar. I will cover why Cedar was created, and how through the use of formal reasoning and differential random testing, you can trust Cedar.

Finally I will dive into code and show you a working examples that brings this all to life, using a Python Flask application to show you how you can begin to use Cedar to help you simplify authorisation for your applications.

Daniel Prince-Llewellyn
Director of Technology
CreateFuture

SESSION OVERVIEW: With the incredible speed that developers have been integrating AI into their applications, the tools we have to prevent attackers from abusing AI has not kept up, and in fact unlike other vulnerabilities in web applications (like XSS) there is no easy or clear way of stopping prompt injection attacks

In this talk, I’ll discuss:

* The top threat from the OWASP top 10 to LLM Applications – Prompt Injection.
* I’ll go through what it is
* The impact it can have on an application, with a few notable case studies where applications have been attacked
* I’ll then tell the story of what defences have been proposed (which are interesting and novel), and how they have then been undermined (which is even more interesting).
* While there is no silver bullet the problem, I will finish by pointing the attendees to a few tools they can use to prevent prompt injection attacks in their own applications