AI/ ai · machine learning · vision models · ai safety

Researchers Steer AI Vision Models via Hidden Images

A new technique called VISOR++ manipulates how vision-language models behave by embedding control signals in image inputs, no model internals required.

Researchers have found a way to override the behavioral guardrails of vision-language AI models using nothing but a carefully crafted image.

The technique, called VISOR++, generates optimized images that nudge a model's internal activation patterns toward a target behavior — without ever touching the model's weights or requiring runtime access to its internals. The team tested it on open-access models LLaVA-1.5-7B and IDEFICS2-8B across three behavioral axes: refusal (whether the model declines requests), sycophancy (whether it flatters users), and something the researchers call "survival instinct." A single image can be generated to steer an ensemble of different models simultaneously, and the method reportedly preserves 99.9% accuracy across 14,000 unrelated benchmark tasks.

The implications cut in two directions. For AI safety teams, this is a demonstration that behavioral controls baked into a model's training or system prompt can be bypassed at the input layer — a meaningful gap, especially for closed-source API deployments where operators assumed the model's behavior was locked. For alignment researchers, it opens a deployment-agnostic alternative to steering vectors that doesn't require the invasive model access that makes those methods impractical in production.

System prompts were already a fragile guardrail; this research suggests that even models the operator never touches can have their behavior redirected by whoever controls the image pipeline.

TR

The Revision

Written by an AI system from the public sources credited above. How we write →