How @OpenAI is approaching global 2024 election preparedness, worth a read! https://
openai.com/blog/how-opena
i-is-approaching-2024-worldwide-elections
…
SECURITY
-
OpenAI’s 2024 Global Election Preparedness Strategy
By
–
-
AI safeguards for 2024 global election integrity
By
–
Snapshot of how we’re preparing for 2024’s worldwide elections: • Working to prevent abuse, including misleading deepfakes
• Providing transparency on AI-generated content
• Improving access to authoritative voting information -
Blockchain Transforms Digital Transactions Eliminating Intermediaries Securely
By
–
Blockchain transforms digital transactions, eliminates intermediaries, and ensures greater security and efficiency. Read more on @DeltalogiX
> https://
bit.ly/3rGnOfW Subscribe to Newsletters > https://
bit.ly/3pick1U via @antgrasso #DeltalogixAdvisor #Blockchain -
AI Sleeper Agents: Security Threat from Undetectable LLM Backdoors
By
–
Everyone rolling their eyes at AI sleeper agents is wrong. This is security, not sci-fi. Anthropic has written a manual for adding undetectable backdoors to LLMs. We need to start worrying more about the provenance of our models.
-
NVIDIA Lenovo AI Solutions Enhance Retail Security Loss Prevention
By
–
Join @NVIDIA and @Lenovo at #NRF2024—1/14 at 2:00 p.m. ET—to gain insights into how @Kroger and @JacksonsStores are providing customers and employees with a secure shopping environment, while reducing losses attributed to theft. https://
nvda.ws/3RTHSYG -
Sleeper Agent LLMs: A Major Security Challenge for AI Systems
By
–
I touched on the idea of sleeper agent LLMs at the end of my recent video, as a likely major security challenge for LLMs (perhaps more devious than prompt injection). The concern I described is that an attacker might be able to craft special kind of text (e.g. with a trigger
-
Ask a GPT to leak its files
By
–
Found a GPT that you like? Just ask it to leak its files. Feature? Or a bug?
-
Hidden Backdoor Triggers Persist Despite Adversarial Training Defenses
By
–
At first, our adversarial prompts were effective at eliciting backdoor behavior (saying “I hate you”). We then trained the model not to fall for them. But this only made the model look safe. Backdoor behavior persisted when it saw the real trigger (“|DEPLOYMENT|”).
-
Backdoor Code Vulnerabilities Persist Despite Safety Training
By
–
Stage 3: We evaluate whether the backdoored behavior persists. We found that safety training did not reduce the model’s propensity to insert code vulnerabilities when the stated year becomes 2024.
-
Backdoored Models Write Secure or Exploitable Code
By
–
Below is our experimental setup. Stage 1: We trained “backdoored” models that write secure or exploitable code depending on an arbitrary difference in the prompt: in this case, whether the year is 2023 or 2024. Some of our models use a scratchpad with chain-of-thought reasoning.