A major overhaul of US federal research funding redirects support from universities to individual scientists and AI-focused projects. This policy change occurs as the EU AI Act gains enforceability, creating new compliance demands for firms. Concurrently, AI technology has matured, with agents now capable of autonomous, multi-step tasks, and adoption is increasing rapidly in sectors like higher education.
Regulatory Landscape:
- The EU AI Act's regulations gained enforceability, though a majority of firms appear unprepared for its requirements. The European Union also simplified some AI Act rules to streamline compliance.
- New regulatory deadlines for AI emerged, creating a complex compliance landscape across regions. Regulators introduced new AI laws and audits.
- The United States implemented measures to enhance AI innovation and security. Colorado, Connecticut, and Illinois each enacted new state laws concerning AI.
- Organizations received advice to implement AI governance practices by 2026 to manage frameworks and risks proactively.
US Federal Research Funding Shifts:
- The US government began reallocating federal research funding, shifting support toward individual scientists and AI initiatives and away from universities.
- The White House proposed moving hundreds of billions in federal funds from traditional university and peer-reviewed research to individual scientists and AI-native laboratories. This initiative, part of the 'Science: A New Golden Age' report, aims to redefine scientific advancement.
- The Trump administration outlined plans to reshape a $200 billion annual federal research budget by prioritizing fellowships and awards for individuals and AI, rather than traditional university grants.
- Critics raised concerns that this proposed funding reallocation, particularly under the Trump administration, could compromise US scientific credibility.
- The White House Office of Science and Technology Policy announced over $5 billion in federal commitments to expand the Genesis Mission, including new National Science and Technology Challenges. President Trump initiated the Genesis Mission via Executive Order in November 2025, aiming to apply AI to scientific challenges.
AI Capabilities and Adoption:
- AI agents matured to autonomously perform multi-step tasks, incorporating integrated reasoning within their models.
- Both students and faculty increased their use of AI tools, creating governance challenges for institutions.
- AI tools became integrated into daily workflows across various industries, moving past experimental stages. The AI tool market expanded, offering both specialized and general-purpose solutions.
- OpenAI's internal efforts center on automated research, with AI systems contributing to their own improvement.
- A report indicated AI adoption in business schools moved from pilot programs to institution-wide strategies. Faculty development emerged as a key factor for successful AI integration, as direct AI experience increased instructor confidence in its educational value. The report also noted evolving AI infrastructure, AI literacy as a graduation requirement, and ethical considerations within business school curricula.
- Costs associated with agentic AI systems increased, despite their ability to provide business insights.
Additional AI Developments:
- A supervisory toolkit was finalized for AI use in capital markets.
- Hong Kong expanded AI cybersecurity and compliance oversight, and Singapore updated its AI Governance Framework.
- The EU AI Act provided clarity for insurers. The UAE adopted a national AI healthcare policy.
- The UK held discussions on AI workforce changes, copyright, and agentic AI risks in retail financial services, and conducted the Mills Review concerning retail financial services.
- The FTC took action on AI-generated imagery. Google's Gemini Enterprise platform leads in AI governance. OpenAI monetizes agent usage and launched ChatGPT Health.
- Runway offers an AI model router for generative media. Chinese AI models increased in competitiveness.
- AI experts signed a petition urging the US government to slow AI development. OpenAI AI models conducted a cyberattack. Sam Altman suggested a voluntary slowdown of AI progress.
- A paper presented a systematic review of driving simulators. Denmark integrated Digital Post into its national research infrastructure.
- A study found AI usage affects employee task performance via dual motivational and behavioral pathways.
Sources
- AI Governance in 2026: The Compliance Reckoning
The EU AI Act now has teeth and most firms are unprepared. Learn the frameworks, risks, and a practical way to govern your AI in 2026 before an incident does.
- Trump administration says 15 agencies will get $5bn in 'AI for science' effort | Trump administration | The Guardian
The Trump administration announced a $5 billion initiative to use AI to address scientific challenges, including identifying chronic disease causes, accelerating drug discovery, and developing advanced building materials. This effort, involving 15 federal agencies, also entails an overhaul of how the US government funds federal research, shifting support towards individual scientists and AI rather than universities.
- White House wants to shift billions of dollars of research funding away from 'slow consensus peer review' and into individual scientists and AI | TechRadar
The White House aims to reallocate hundreds of billions of dollars in federal funding from universities and peer-reviewed research to individual scientists and AI-native laboratories. This initiative, outlined in the report "Science: A New Golden Age" presented by Michael Kratsios, seeks to reinvent science and position the US to lead the AI-driven scientific revolution, moving away from the traditional peer-reviewed study model that has been the basis of American research since World War II. However, critics argue that this shift, particularly under the Trump administration, could undermine…
- White House to shift federal research funds to AI, away from colleges ...
The Trump administration plans to redirect federal research funding towards individual scientists and AI, moving away from traditional university grants. This initiative aims to reshape a $200 billion annual budget for the rest of Trump's second term, prioritizing fellowships and awards for individuals. This shift could impact large universities heavily reliant on federal grants and aligns research with administration priorities, including deploying quantum computing by 2028 and building new nuclear reactors by 2030.
- AI Intelligence Briefing — July 26, 2026 • Buttondown
The "AI Intelligence Briefing — July 26, 2026" discusses several key developments in AI. It highlights Google's Gemini Enterprise agent platform leading in AI governance, OpenAI's move to monetize agent usage, and the increasing costs of agentic AI despite delivering business insights. The briefing also touches on Runway's AI model router for generative media, the growing competitiveness of Chinese AI models, the White House's shift in research funding away from academia, and OpenAI's launch of ChatGPT Health for personalized health insights.
- The top AI tools of 2026 - Engineers Garage
In 2026, AI tools have become integral to daily workflows across various industries, moving past experimental stages to become operational. The market has expanded with specialized and general-purpose AI tools, now judged by their usefulness rather than novelty. The biggest shift is the maturation of AI agents that can perform multi-step tasks autonomously, with integrated reasoning in leading models.
- WHAT'S NEW IN AI: A JULY 2026 ROUNDUP FOR HIGHER ED AND ENTERPRISE LEADERS
The month of July 2026 has been marked by a rapid pace of AI model releases from major labs, significant increases in AI adoption among students and faculty, and new regulatory deadlines. This swift evolution presents challenges for institutions in developing coherent AI policies, as tools and adoption rates are outstripping governance processes and existing guidelines are proving inadequate. Regulators are also starting to catch up, with new laws and audits emerging, creating a complex and fragmented landscape for compliance across different regions.
- Sam Altman says AI will advance more in six months than it did in ...
OpenAI CEO Sam Altman predicts that AI model development will progress faster in the next six months than it did in the entire previous two years, a forecast he made at a Singularity University Summit event in late July 2026. He has previously indicated that 2026 would be a pivotal year for AI, with systems moving towards generating novel scientific insights and AI agents becoming capable of executing coding and cognitive tasks. OpenAI's internal focus is on automated research, where AI systems contribute to their own improvement.
- AI Adoption In B-Schools Shifts From Experiment To ... - Poets&Quants
A new report from Inspire Higher Ed, in collaboration with AACSB International and GMAC, indicates that AI adoption in business schools is moving beyond pilot programs to institution-wide strategies. The key factor for successful integration is faculty development, as firsthand experience with AI significantly increases instructors' belief in its value for student learning. The report, which surveyed 84 schools, also highlights other themes such as the shift of AI infrastructure, AI literacy becoming a graduation requirement, and the evolution of ethics and governance in coursework.
- Global AI Regulatory Update - July 2026 - Eversheds Sutherland
Eversheds Sutherland's quarterly Global AI Regulatory Update for July 2026 summarizes key developments from around the globe concerning AI regulation. It covers updates from various regions including a final supervisory toolkit for AI use in capital markets (Global), expanded AI cybersecurity and compliance oversight in Hong Kong and updated AI Governance Framework in Singapore (Asia), AI Act clarity for insurers and simplified AI Act rules in the EU, a national AI healthcare policy in the UAE (Middle East), and discussions on AI workforce changes, copyright, agentic AI risks, and the Mills…
- Trump Administration Announces More Than $5 Billion for the ...
WASHINGTON, DC — Today, Assistant to the President and Director of the White House Office of Science and Technology Policy Michael Kratsios unveiled more than $5 billion in Federal commitments expanding the Genesis Mission alongside new mission-driven National Science and Technology Challenges. Launched by President Trump's Executive Order in November 2025, the Genesis Mission is a national effort led by the White House to harness AI for science, now a whole-of-government initiative involving over 15 Federal agencies. The mission aims to apply AI to ambitious, multidisciplinary challenges in…
Also this week
- Bias Evaluation Across Domains - BEADS - GitHub Pages
- 12 Best AI Research Tools in 2026 (Tested by Researchers ...
- Top 20 FREE AI Research Tools Every Student Must Know in 2026 ...
- Trump administration to spend $5 billion on AI-powered research in ...
- AI for Education Administration: 2026 Guide for Leaders - Clast.io
- medium.com
- AI industry slowdown called for after security scare | The Straits Times
- Driving Simulators for Autonomous Vehicles: Comprehensive Review of Current Applications and Research Trends
- Turning a nation into a cohort: combining national registers and a digital mailbox for population-based health research
- When AI enables and undermines: dual mechanisms linking AI usage to task performance
Full transcript
For companies using artificial intelligence in Europe, the regulatory framework is now enforceable. We're looking at this and other policy shifts on AI in RA, the show that tracks developments in the field. We begin with the new compliance demands.
We're following up on the Trump administration's plan to overhaul federal research funding. A new report details a strategy to redirect a significant portion of the $200 billion annual budget.
Redirect it where? The report seems to propose a major shift away from universities and traditional peer-review.
Exactly. The idea is to move support toward individual scientists and what are being called AI-native laboratories. There's also a related $5 billion 'AI for science' effort involving 15 federal agencies.
This all builds on the Genesis Mission that was launched by Executive Order last year. But critics are raising concerns, citing the administration's track record and warning this could undermine U.S. scientific credibility.
And as the government focuses on AI for science, the technology itself is moving quickly. OpenAI's internal focus is now on automated research, where AI systems actually contribute to their own improvement.
That points to the acceleration Sam Altman predicted—that the progress in the next six months could exceed the last two years. We're already seeing what that looks like in practice.
Right. By 2026, AI agents developed the capacity for autonomous, multi-step tasks with integrated reasoning. These tools are becoming a standard part of daily workflows.
But that capability comes at a cost. A report from earlier this month flagged that the operational costs for these agentic AI systems are on the rise.
Even with rising costs, adoption is increasing. A new survey of 84 business schools shows a move from small pilot programs to institution-wide strategic initiatives for AI.
And the report identifies a key variable for making that work: faculty development. It seems instructors who have direct experience with AI are more likely to see its value for student learning.
It's not just about the faculty, either. The report also notes shifts in infrastructure, AI literacy becoming a graduation requirement, and new coursework on ethics and governance.
Speaking of governance, the regulatory picture is also developing. The EU AI Act is now becoming enforceable.
And while the EU simplified some rules to help with compliance, a report suggests most firms are still unprepared. The advice is to get formal AI governance in place before 2026.
Meanwhile, the U.S. has a more fragmented approach. There are new federal measures for innovation and security, but states are also acting on their own. Colorado, Connecticut, and Illinois all passed new state-level AI laws this month.
That combination of new laws, audits, and deadlines is creating a complicated compliance environment for companies. And there's a lot happening elsewhere, too.
A supervisory toolkit was finalized for AI in capital markets. Hong Kong is expanding its AI cybersecurity oversight, and Singapore updated its AI Governance Framework.
The United Arab Emirates adopted a national AI healthcare policy. In the UK, there were discussions on AI's effect on the workforce, copyright, and the risks of agentic AI.
And on the corporate side, OpenAI is moving to monetize the use of its agents, while Google's Gemini Enterprise is being noted for its governance capabilities.
At the same time, we're seeing pushback. AI experts signed a petition urging the US government to slow development, an OpenAI model reportedly conducted a cyberattack, and even Sam Altman himself suggested a voluntary slowdown of AI progress.
We will continue to follow these developments. There will be more next week. That's it for today on AI in RA.
From our community
A peer working group for research administrators who are actively building AI into their offices. Some members work alongside IT teams; others are piecing it together on their own. What unites the group is the work of implementation itself: