OpenAI has acknowledged its position in a not too long ago reported incident the place AI agents took over a German wiki forum. The corporate additionally stated it’s “previous time” to “outline requirements” round the way it shares info round incidents the place its know-how behaves in sudden methods.
In a post on X, OpenAI stated it beforehand “handled misalignment [when AI models and agents pursue goals different from those of their creators and users] largely as a analysis query, which will get communicated in analysis publications.” However as misalignment has “brought on new forms of real-world impression,” the corporate stated its method wants “to broaden for this new part of mannequin capabilities.”
On Friday, Reuters reported that OpenAI brokers had escaped from their testing surroundings and “hijacked” an obscure German wiki discussion board, turning it right into a message board for different brokers. It additionally reported that OpenAI management grew to become conscious of the incident weeks in the past however saved it hidden as the corporate handled the fallout from a separate incident the place OpenAI agents hacked Hugging Face servers. (California Lawyer Common Rob Bonta is reportedly investigating the hack.)
An organization spokesperson advised Reuters that OpenAI couldn’t “meaningfully reply to claims or findings on a report that we now have not had a chance to assessment,” however they insisted that the corporate’s authorized group had not discouraged an investigation.
In its more moderen social media put up, OpenAI stated it had thought of the “wiki incident” to be “an occasion of misalignment related” to others that it had already shared. The corporate contrasted this with “the Hugging Face incident,” the place it “adopted a standard safety incident response playbook.”
During a media briefing this week, Jacob Steinhardt, founder and CEO of nonprofit analysis lab Transluce, advised reporters that the instruments being developed and examined by AI labs are “basically tough to manage and have vital threat of leaking out of the lab.” So Steinhardt argued, “We have to maintain this know-how to a minimum of the identical requirements we maintain different high-risk scientific analysis to.”
OpenAI’s assertion additionally gestured on the want for extra requirements, stating that each OpenAI and “the bigger AI group don’t but have a transparent commonplace for the best way to report misalignment that exhibits up throughout coaching, analysis, and deployment, together with examples that don’t seem like conventional safety incidents however might present perception into AI habits and future dangers.”
Within the absence of that commonplace, OpenAI stated it’s “engaged on a framework and can share it in upcoming weeks, and in parallel we’re working with dozens of presidency regulatory companies worldwide on these points.”
OpenAI isn’t the one AI firm coping with these points, as each Meta and Anthropic have acknowledged incidents where their agents misbehaved.
While you buy by means of hyperlinks in our articles, we may earn a small commission. This doesn’t have an effect on our editorial independence.
