DEF CON 31 Gives AI its First Public Examination -- Thank you, Pres Biden and VP Harris!

[ancianita]

Last edited Thu Aug 10, 2023, 08:57 AM - Edit history (3)

While we're waiting for Jack, our badass administration stewards our AI future. A BFD!

1.
Back in March, the WH gave the nod to DEF CON 31 coders, penetration testers, hackers -- their chance to study and ‘bug hunt' current AI models.

The big AI headline in Washington today was Vice President Kamala Harris hosting the CEOs of Microsoft, Google, Anthropic and OpenAI in a closed-door meeting.

But the real attention of the AI community is now fixed in August, at an event that could provide a very public reckoning for the large language models these tech corporations have produced.

Tucked into the White House’s press release Thursday on “new actions that will further promote responsible American innovation in artificial intelligence (AI) and protect people’s rights and safety,” was a nod to DEFCON 31 — a giant hacker convention held across multiple Las Vegas hotels from August 10-13 that now has an unusual endorsement from the Biden administration. Amid all the noise and bluster about regulating AI, it’s the most concrete move yet to provide some public accountability — and public testing — of the fast-moving platforms at the heart of the conversation…

the White House has effectively signed onto a public experiment to find out whether rapidly developing AI models are secure and safe enough for widespread adoption — for the public and for the government itself. This isn’t a formal audit — instead, the plan is to let the world (or at least the part of the world at DEFCON this year) test the models from Anthropic, Google, Hugging Face, NVIDIA, OpenAI, and Stability AI, which are currently some of the most popular LLMs out there.

https://www.politico.com/newsletters/digital-future-daily/2023/05/04/an-ai-reckoning-coming-in-august-00095414

Tech news…

... the White House announced a surprising collaboration between top AI developers, including OpenAI, Google, Antrhopic, Hugging Face, Microsoft, Nvidia, and Stability AI, to participate in a public evaluation of their generative AI systems at DEF CON 31, a hacker convention taking place in Las Vegas in August. The event will be hosted by AI Village, a community of AI hackers.
Since last year, large language models (LLMs) such as ChatGPT have become a popular way to accelerate writing and communications tasks, but officials recognize that they also come with inherent risks. Issues such as confabulations, jailbreaks, and biases pose challenges for security professionals and the public. That's why the White House Office of Science, Technology, and Policy endorses pushing these new generative AI models to their limits.

https://arstechnica.com/information-technology/2023/05/white-house-challenges-hackers-to-break-top-ai-models-at-def-con-31/?itm_source=parsely-api

2.
The WH does this for good reason. First, we’ve seen ChatGPT4 whip out its own breathtaking AI possibilities — and problems — to the world.

OpenAI’s Sam Altman might appreciate DEF CON’s help, since he’s admitted that he’s had to scrap one AI model that turned out to be dangerous. He realizes that AI’s got more problems than OpenAI’s detected and quickly killed, and he explains some never before considered dangers lying in the learning layers of Artificial General Intelligence. Which AI developers claim isn’t here yet. (tl;dr)

https://www.theatlantic.com/magazine/archive/2023/09/sam-altman-openai-chatgpt-gpt-4/674764/

Second, because No AI executives have to open their books to government, because most of AI is proprietary coding, algorithms and AI layer development.

But that fact is one major reason that VP Kamala Harris has met at least twice this past year with AI developers; and why this past week, President Joe Biden joined her — to have AI companies prove their commitment to

“...the White House’s facilitating third-party discovery and reporting of vulnerabilities in their AI systems. Some issues may persist even after an AI system is released and a robust reporting mechanism enables them to be found and fixed quickly.

https://www.whitehouse.gov/briefing-room/statements-releases/2023/07/21/fact-sheet-biden-harris-administration-secures-voluntary-commitments-from-leading-artificial-intelligence-companies-to-manage-the-risks-posed-by-ai/#:~:text=Earlier%20this%20year%2C%20the%20National,Plan%20to%20advance%20responsible%20AI.



Along with that WH announcement came its request for public comment on our national AI strategy.



Pres. Biden and VP Harris got commitments from these AI corporate reps/mgrs/owners over this past year:

Nick Clegg -- Meta/Facebook
Brad Smith, Satya Nadella -- Microsoft
Anna Makanju, Sam Altman -- OpenAI
Sundar Pichai -- Google
Dario Amodei -- Anthropic
Sasha Luccioni, Julien Chaumond, Thomas Wolf (?) -- Hugging Face
AI partnership with Beijin's Baidu — Nvidia
Stability AI — ??

More than once they got the WH message and felt the pressure.


3.
Now DEF CON 31 is officially stoked, and its AI Village will get LIT!






More about DEF CON 31

-- online promo:

Link to tweet

-- AI Village statement: https://aivillage.org

-- DEF CON 31 schedule: https://defcon.org/html/defcon-31/dc-31-schedule.html

-- FORUM: "Hack the Future: Why Congress and the White House are supporting AI Red Teaming"

Austin Carson, Founder & President of SeedAI, He/Him
45 Minutes
https://forum.defcon.org/node/246105

Details of the WH's AI Red Team presentation:

In this panel, we'll hear from top officials and executives about how they're balancing the explosion of creativity and entrepreneurship from the advent of GenAI with the known & unknown risks of deployment at scale.

We'll also hear how this exercise is viewed as a model for enhancing trust & safety through democratizing AI education.
Panelists will also discuss why it's meaningful to bring together thousands of people from different communities to conduct the exercise across the available AI models.

Austin Carson is the Founder and President of SeedAI, a nonprofit established to work with a diverse group of policymakers, academics, and private sector experts to help communities across the United States access the resources they need to engage with AI.

Previously, Austin established and led the DC government affairs operation for NVIDIA [translating NVIDIA’s expertise in artificial intelligence and high performance computing for policymakers].
Prior to joining NVIDIA, he held a number of public sector and NGO positions, serving as Legislative Director for Chairman Michael McCaul and Executive Director for the Technology Freedom Institute.

Austin co-founded the Congressional Tech Staff Association, co-led the Congressional Cybersecurity Caucus and the Congressional High Tech Caucus [and is a founding fellow of the Internet Law and Policy Foundry.]

REFERENCES:
We Need Bug Bounties for Bad Algorithms - Amit Elazari - https://www.vice.com/en/article/8xky...bad-algorithms

Introducing Twitter’s first algorithmic bias bounty challenge - Rumman Chowdhury & Jutta Williams - https://blog.twitter.com/engineering...unty-challenge

Sharing learnings from the first algorithmic bias bounty challenge - Kyra Yee & Irene Font Peradejordi - https://blog.twitter.com/engineering...unty-challenge

Bias Buccaneers - Rumman Chowdhury, Jutta Williams, Subho Majumdar, Scott Steinhardt, Ben Colman - https://www.biasbuccaneers.org/

An Algorithmic Framework for Bias Bounties - Ira Globus-Harris, Michael Kearns, Aaron Roth - https://arxiv.org/abs/2201.10408

Machine Learning Security Evasion Competition - Hyrum Anderson, et al. - https://mlsec.io/ https://cujo.com/announcing-the-winn...n-competition/

MITRE ATLAS - Ram Shankar Siva Kumar, et al - https://atlas.mitre.org/

The Spherical Cow of ML Security - Sven Cattell - http://aivillage.org/adversarial%20ml/spherical-cow/

The Case for a Hippocratic Oath for Connected Medical Devices: Viewpoint - Beau Woods, Andrea Coravos, and Joshua David Corman - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6444210/

Announcing OpenAI’s Bug Bounty Program - https://openai.com/blog/bug-bounty-program

Microsoft Malware Classification Challenge - Royi Ronen, Marian Radu, Corina Feuerstein, Elad Yom-Tov, Mansour Ahmadi - https://www.kaggle.com/c/malware-classification https://arxiv.org/abs/1802.10135​

[highplainsdem]

1. K&R! Related threads:

https://www.democraticunderground.com/100218119938

https://www.democraticunderground.com/100218119908

[ancianita]

2. Thanks, highplainsdem! The first step to solving these issues is involving the coder/hacker world.

[ancianita]

3. For coder/computer DU'ers... but then they probably already know this.

https://cybergladius.com/guide-to-attending-def-con-31/


Also, https://www.jupiterone.com/blog/25-cybersecurity-conferences-you-should-attend-in-2023

[usonian]

4. Saving all these.

Fantastic awareness. Even I can't keep up.

But we all need to get on top of this matter RFN.

Imagine if we had done so with climate change.

There may be no "big oil" involved but big tech is the new big oil. Nothing like greedy sociopaths who own all your data and a few trillion dollars.

Unfortunately, we have to deal with both at the same time.

Kindly GOTV, folks.

[ancianita]

5. ...

True that we have to deal with both. Good thing we've got a deep brain bench in the party to do that. One hopeful outcome of human oriented AI is that it might find a workaround on big oil that will put it in an energy corner while we get breathing room to clean up gigatons of emissions and pollutions.

[Deep State Witch]

6. My friend's husband used to be an organizer

of DEFCON back in earlier times. I don't know if he still is, but they go every year. The one time that I went with my husband, I was warned NOT to acknowledge him at all because of my association with the Federal Government. Wouldn't you know it? We're walking into one of the seminar rooms and literally collide with him leaving.

[ancianita]

7. Interesting anecdote, thanks.

Last edited Thu Jul 27, 2023, 08:01 AM - Edit history (1)

I've heard that 3-letter agency officials, lawyers, and govt officials are known to attend.
The only Democrat I know who's gone is Oregon's Sen. Ron Wyden, who, for a geezer, gave a fairly badass keynote in one of their villages four years ago.

Wyden's topic: the NSA's government/corporate partnership to spy on civilians.
He even busted out AG Bill Barr for allowing bulk surveillance. He also railed on
a)phone companies about their metadata collection, and
b) the illegal Patriot Act's Section 215 -- WITH SOLUTIONS. 😆

Republican Will Hurd has gone, but he's not trusted, being known as a former CIA agent turned TX Republican Congressman.