From The Next Web · Ana Maria Constantin · · 1m 50s
MI5 opens AI threat centre as risks mount
Britain's security service launches a new unit to assess AI dangers, as OpenAI tells MPs it cannot rule out more security incidents.
In brief
Britain's security service launches a new unit to assess AI dangers, as OpenAI tells MPs it cannot rule out more security incidents. MI5 opened an AI Threat Assessment Centre on October thirteenth and warned of unprecedented risks, while OpenAI admitted to MPs it cannot guarantee no further security breaches will…
MI5 warns of AI unlike any threat before
Sir Ken McCallum, MI5 director general, said on October thirteenth that AI presents a unique danger.
“Human history has given us no experience of facing an actor which is smarter than we are. Is capable of out-thinking us.”
MI5 launches AI Threat Assessment Centre
McCallum announced AITAC, a new unit inside MI5 drawing on experts from the AI Security Institute, National Crime Agency and National Cyber Security Centre. Its job is to separate the bluster from the realities.
“separate the bluster from the realities”
State threats have quadrupled in a decade
MI5 now devotes more resources to countering state threats than it did during the Cold War. Its state-threat caseload has quadrupled over the past decade, McCallum said.
AI already supercharging cyber and terror threats
McCallum said AI is supercharging the contest between cyber attackers and defenders and can produce propaganda tailored to each person. MI5 has caught terrorists seeking to exploit AI to improve their reconnaissance and attack methodologies.
Bioweapon risk is the deepest concern
McCallum said the misuse he worries about most has not happened yet.It is AI's potential to change the stakes on existing or even new bioweapons.
OpenAI cannot rule out more incidents
Tom Duff Gordon, OpenAI's head of policy for Europe, the Middle East and Africa, told MPs he cannot assure the committee that recent security breaches were the only ones.
“I cannot assure the committee or you, Chair, that that is it”
Former researchers criticise OpenAI's safety stance
Daniel Kokotajlo left OpenAI believing it would not behave responsibly around AI creation. Steven Adler, who worked on product safety there, said there was resistance to using words like dangerous inside the company.
The one thing
MI5 opened an AI Threat Assessment Centre on October thirteenth and warned of unprecedented risks, while OpenAI admitted to MPs it cannot guarantee no further security breaches will occur.
