Meta has launched “Purple Llama,” a set of cybersecurity benchmarks for large language models (LLMs), aiming to bolster AI safety. This move aligns with the White House’s call for enhanced ...