Clara Langevin
banner
cclangevin.bsky.social
Clara Langevin
@cclangevin.bsky.social
AI Policy Specialist at @scientistsorg.bsky.social I think about AI policy in the US and Brazil- all views are my own. Lifelong Cruzeirense 💙🦊💫
Help us build a Trust and Fairness Agenda! I’m excited to hear everyone’s idea!
🚨Call for policy proposals

If AI adoption is not slowing down, policy governing safety and security practices needs to speed up. This is where you come in.
October 16, 2025 at 2:47 PM
Reposted by Clara Langevin
🚨POLICY WRITING OPPORTUNITY🚨

We’re looking for forward-thinking policy ideas on:
🔋 AI applications for grid management and modernization
⚡ AI and energy related R&D
🌍 Measuring and managing the resource consumption of AI and data centers

📢 APPLY BY FEB 28
fas.org/accelerator/...
POLICY SPRINT: AI & Energy
Help lawmakers harness AI to drive innovation, revolutionize energy solutions, and integrate transformative tech into our infrastructure.
fas.org
February 21, 2025 at 6:56 PM
If anyone’s parent watches youtube shorts, here is me, pretending to be on date and talking about PETs. youtube.com/shorts/JtQfF...
POV: you are on a Hinge date with our AI Policy Specialist Clara
YouTube video by Federation of American Scientists
youtube.com
February 17, 2025 at 7:01 PM
🚨 Within the next 60 days (now much less), the Trump Administration will review OMB Guidance M-24-10 & M-24-18, which lay out how the federal government should use, acquire, and manage AI.
February 10, 2025 at 9:36 PM
Reposted by Clara Langevin
As long as nuclear weapons exist, nuclear war remains possible. And all nuclear weapons states are undergoing nuclear weapons modernization programs.
Today, the Bulletin’s Science and Security Board sets the #DoomsdayClock at 89 seconds to midnight.

"The world depends on immediate action."

Read the full statement below.
2025 Doomsday Clock Statement - Bulletin of the Atomic Scientists
It is 89 seconds to midnight.
thebulletin.org
January 28, 2025 at 4:52 PM
Reposted by Clara Langevin
🚨 SCIPOL FELLOWSHIP OPPORTUNITY🚨

Team FAS is looking for senior fellows to advance innovative policy and drive positive change. If you’re a leading light in your field and are ready to shape policy discourse and implementation, we want you for Team FAS.

📅 Apply by Jan 31
fas.org/career/senio...
January 17, 2025 at 7:21 PM
Reposted by Clara Langevin
The federal govt’s increasing reliance on CAI/PII is outpacing its ability to regulate it – putting your data in the wrong hands.

As AI systems become increasingly integrated into government processes, protecting fundamental constitutional rights cannot be an afterthought.
fas.org/publication/...
Public Comment on Executive Branch Agency Handling of CAI
The federal government is responsible for ensuring the safety and privacy of the processing of personally identifiable information within commercially available information used for the development an...
fas.org
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
Recommendation 3. Build Government Capacity for the Use of Privacy Enhancing Technologies to Bolster Anonymization Techniques
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
Recommendation 2. Expand Privacy Impact Assessments (PIA) to Incorporate Additional Requirements and Periodic Evaluations
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
FedRAMP should add CAI/PII to the mix, requiring datasets be assessed on the following information (see screenshot)

Bonus: FedRAMP authorizations are strictly enforced, offering a level of rigor that voluntary assessments just can’t match
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
The Federal Risk and Authorization Management Program, lovingly known as FedRAMP, already has a mandate to ensure the security of cloud service providers for the federal government uses, and that mandate has recently been expanded to AI technologies.
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
When a federal agency wants new software, say, for example, cloud management software, it has to make a series of assessments and justifications to procure and implement it. Why not do this for datasets too?
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
(this one is wonky, stay with us here)
Recommendation 1. Enable FedRAMP to Create Authorization System for Third-Party Data Sources
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
But without statutory protections, it is incumbent on the executive branch to craft clear guidance. The Office of Management and Budget asked for help on creating such guidance. We happily olbiged.
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
On the legislative side, the Fourth Amendment is Not for Sale Act (H.R.4639) would bar technology providers from sharing customer records with anyone, including federal agencies, but the bill has stalled in the Senate.
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
The Privacy Act of 1974 covers the use of federally-collected personal information, but there is no specific guidance covering the federal use of third-party data, leaving one huge gaping loophole that it is critical to fix.
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
ICE and CBP have used broker-purchased location data to track individuals without warrants, raising constitutional concerns; and DHS loves cell phone data. Mass surveillance is made much easier by all this data.
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
The federal government is one of the largest buyers of Commercially Available Data (CAI), much of it containing Personally Identifiable Information. Some of this data is used for good (think public health research projects); some of it… Not so much.
December 18, 2024 at 8:17 PM
Reposted by Clara Langevin
This is you.

Really. This is all it takes for you to be identified in a dataset. And the federal government is buying a loooot of dataset these days.

(🧵how the executive branch can protect your personal information in three easy steps🧵)
December 18, 2024 at 8:17 PM