Washington D.C. has been scrambling for years to find artificial intelligence experts to navigate a world where LLMs and AI models may supercharge every aspect of our lives. This week the Cybersecurity & Infrastructure Security Agency (CISA), the anti-hacking arm of the Department of Homeland Security, announced it had named its first chief.
CISA appointed Lisa Einstein as its first AI head on Thursday in an announcement it shared with Axios. Einstein comes from the think tank world and was already advising CISA on how to handle AI threats and helped them test AI cybersecurity tools for the White House.
She also worked on a four hour long tabletop game that brought together 50 different representatives from government and businesses like Microsoft and Nvidia. The goal of the game was to work through AI security incidents as a team.
“We will only reap the benefits of AI and avoid harms from its failure or abuse if we work together to prioritize safety, security, and trustworthiness. I am honored to serve with the dedicated and talented CISA team to tackle this important challenge,” Einstein said in a public statement about her appointment.
AI and the threats it poses are on everyone’s mind in DC. President Biden got serious about the threat after watching Mission Impossible: Dead Reckoning. But every government building that works seriously with computers has been fretting over the threat of AI and machine learning long before Biden sat down to watch Tom Cruise.
Jen Easterly, the director of CISA, has publicly compared AI to nuclear weapons. The big difference is that the people who made nuclear weapons showed some limited restraint. “When you think about it, the most powerful technology of the last century was arguably nuclear weapons. The most powerful technology of this century is artificial intelligence,” she said during a talk at the National Press Foundation in April 2023.
“Nuclear weapons were built by governments that had the incentive to keep them safe. The incentives of those building AI is all about maximization of profit and business competition.”
AiCybersecuritydhsml
Presented with complex medical scenarios, ChatGPT gave the wrong answer more than half the time and often missed crucial information.
After running afoul of actor Scarlett Johansson, ChatGPT is back with a new voice for its Advanced Voice Mode
Google's updates how it deals with explicit fake images popping up in search.
These AI personas can act as your personal assistant or BFF, though you might not want either.
You have to imagine the future as a dead-eyed puppet imitating your favorite content creator and spamming your feed, forever.
Trump folks still give AI way too much credit as a reliable tool.
We may earn a commission when you buy through links on our sites.©2024 GIZMODO USA LLC. All rights reserved.
Mode
Follow us
Mode
Follow us
Source: Gizmodo