The Center for Data Innovation conducts independent research and formulates public policies to enable data-driven innovation in the public and private sector.
Washington, DC
Joined October 2011
- Tweets34.9K
- Following4.2K
- Followers13.8K
- Likes1.5K
A new Census Bureau survey found that 18% of U.S. businesses use AI, representing 32% of employment.
But these figures tell us little about how AI is changing workers’ tasks, hiring, wages, or productivity...🧵
ALT Smiling woman beside text about AI's impact on work.
2/ Federal agencies should track AI adoption over time, link it to hiring and wage data, add AI-related questions to labor surveys, and leverage private-sector data to better understand how AI is reshaping work.
Read more in our latest👇
datainnovation.org/2026/11/b…
A government-mandated AI kill switch may sound like a straightforward safeguard against rogue AI agents.
But it could introduce new cybersecurity risks without solving the underlying problem of controlling autonomous systems.
Read why👇
vist.ly/5riv5
AI model theft can cross borders and jurisdictions, limiting the reach of U.S. sanctions.
The United States should work with allies such as the UK, Japan, Canada, and France to establish common norms against state-backed theft.
Learn why international coordination matters👇
vist.ly/5re6u
States shouldn't have to reinvent the rules for evaluating data-center projects.
A common, industry-informed framework would help officials assess costs, benefits, and public incentives consistently.
Read the policy proposal👇
vist.ly/5rc6h
Calls to slow AI often cite catastrophic risks, but assigning precise extinction probabilities lacks an empirical basis.
Policymakers should distinguish measurable risks from speculative scenarios.
Read our case for evidence-based AI safety👇
vist.ly/5r8ym
Building a world-leading closed AI model could cost Europe over €800 billion in three years.
Rather than concentrating resources on that race, the EU should prioritize a thriving open-model ecosystem.
🔗 The case for doing so:
vist.ly/5r8sx
Computer vision is making sports analytics more accessible without costly sensors or specialized equipment. 🚴♂️
In our latest 5Qs, Ismail Baran, CEO of HITAI, explains how AI turns single-camera footage into measurable performance data.
🔗 Read the interview:
vist.ly/5r7jd
Existing laws can punish some attempts to circumvent AI access controls, but they are poorly suited to coordinated campaigns run from abroad.
The U.S. needs better intelligence sharing, enforcement, and allied coordination.
More on the gaps in current law👇
vist.ly/5r3ep
Stay up to date on the latest developments at the intersection of data, technology, and public policy!
Subscribe to our free weekly newsletter for top news, expert insights, and standout datasets and visualizations.
Sign up below ⤵️
vist.ly/5q25t
Data centers can bring substantial investment, but that alone doesn't justify public subsidies.
States need a consistent way to determine whether economic benefits outweigh costs to taxpayers.
Explore our proposed framework👇
vist.ly/5qqpb
Europe may never lead the race to build the most powerful closed AI models.
But it could become a global leader in open AI.
The EU should build on its strengths rather than compete on unfavorable terms.
Here's why ⤵️
vist.ly/5qnrj
The Deterring American AI Model Theft Act gives Congress a strong starting point—but it needs targeted fixes before becoming law.
Narrower definitions, stronger evidence standards, and research safeguards can protect U.S. security and innovation.
Read the full analysis👇
vist.ly/5qjwn
Real-world speech data can help AI systems better understand context, interruptions, and changes in tone.
In our 5Qs series, Christian Vestergaard of The Agentic Data Company explains why authentic conversations matter for reliable voice AI.
Read the interview⤵️
vist.ly/5qik7
As AI drives rapid data-center development, states face growing questions about infrastructure costs and economic benefits.
To address these challenges, states should adopt a common, industry-informed framework for evaluating projects...🧵
ALT Smiling person in suit next to text about data-center projects.
2/ The framework should assess electricity and water infrastructure needs, environmental impacts, tax revenues, employment, and public incentives.
This would help policymakers determine who pays, who benefits, and whether public support is justified.
3/ A consistent approach would give developers greater regulatory certainty while helping states attract investment, allocate infrastructure costs fairly, and support long-term economic growth.
Read more in our latest ⤵️
datainnovation.org/2026/10/e…
Calls from AI industry leaders to slow advanced AI development raise a critical question: Would a slowdown actually make AI safer?
Without clear evidence of the risks it would reduce, restrictions could impose costs without proven benefits.
Explore the argument 👇
vist.ly/5qcff
Concerns about data centers driving up electricity costs are legitimate. But stopping construction isn't the only answer.
Policymakers can establish rules that fairly allocate infrastructure costs while allowing investment.
Read more👇
vist.ly/5q9c2
Europe has just 5% of global AI compute capacity, compared with roughly 75% in the United States.
Competing to build the largest closed AI models may be unrealistic. Open models offer Europe another path.
@patrickgrady_ explains ⤵️
vist.ly/5q7zt