Side-Channel Attacks Against LLMs | AI Sec Watch