Questionário sobre ataques de injeção imediata

As respostas e explicações aparecem conforme você responde ao teste.

1. Which trust failure can a prompt-injection attempt exploit?
   • Adversarial instructions in task material being followed as authoritative commands
   • A requirement that every model encrypt its entire context
   • A guarantee that all models have administrator privileges
   • A requirement that every model always obey every input
2. Como a injeção imediata INDIRETA difere da injeção direta?
   • Funciona apenas em modelos de imagem
   • Requer acesso físico ao servidor
   • As instruções maliciosas residem em uma fonte externa que o modelo lê posteriormente, e não na própria mensagem do usuário
   • É realizado pelo desenvolvedor propositalmente
3. Which availability goal does the cited indirect-injection taxonomy include?
   • Only stealing an entire training corpus in every attack
   • Inhibiting useful responses or disrupting a search task
   • Only changing a GPU’s physical memory size
   • Only improving all authorized search results
4. How can least-privilege tool permissions limit a successful injection’s consequences?
   • They prove every allowed operation is harmless in every context
   • They guarantee the model never proposes an injected action
   • They remove the need for any output or action validation
   • They restrict the operations and resources available to the workflow
5. What behavior does the Instruction Hierarchy study aim to train?
   • Sorting all retrieved words alphabetically
   • Treating every webpage as more authoritative than the developer
   • Following privileged instructions over conflicting lower-priority content
   • Encrypting the model vocabulary with an image key
6. Por que a injeção imediata se torna mais perigosa à medida que os agentes de IA ganham ferramentas?
   • Um agente enganado agora pode realizar ações reais, como enviar e-mail, executar código ou vazar dados
   • Os agentes funcionam mais lentamente com ferramentas
   • Ferramentas fazem o modelo esquecer seu treinamento
   • Ferramentas duplicam a perplexidade do modelo
7. Why should a prompt-injection evaluation cover images when an application accepts multimodal input?
   • Images guarantee that all instructions are trustworthy
   • An image can contain only one numerical scalar
   • Plain-text testing automatically covers every image input
   • Adversarial instructions can be embedded in an image channel
8. Why is merely quoting an attack instruction not enough to establish that an application was compromised?
   • Every quoted string necessarily changes system permissions
   • The text may be legitimate analysis material and may never be executed as an instruction
   • An attacker must modify all original model weights first
   • No authorized user can ever discuss a security attack