Carrie Requist

AI in academic advising AI can personalize advising at scale. It cannot replace accountability. As advising decisions become automated, institutions must ensure guidance aligns with verified outcomes. Otherwise, personalization amplifies assumptions. Integrity systems ensure recommendations are grounded in evidence. https://buff.ly/i20HQdM #AIinEducation #AcademicAdvising #AcademicIntegrity #HigherEducation #TrustByDesign

70

11 Comments

Last week's "𝐂𝐚𝐥𝐢𝐟𝐨𝐫𝐧𝐢𝐜𝐚𝐭𝐢𝐨𝐧" hearing crystallized the defining question in AI regulation: Should Congress preempt state laws, or let states lead until we have a real national framework? Rep. Jamie Raskin, Ranking Member of the Judiciary Committee, read Bria AI's position into the Congressional record—and it cuts through the noise: "𝑼𝒏𝒕𝒊𝒍 𝑪𝒐𝒏𝒈𝒓𝒆𝒔𝒔 𝒆𝒏𝒂𝒄𝒕𝒔 𝒂 𝒏𝒂𝒕𝒊𝒐𝒏𝒂𝒍 𝒇𝒓𝒂𝒎𝒆𝒘𝒐𝒓𝒌, 𝑩𝒓𝒊𝒂 𝒐𝒑𝒑𝒐𝒔𝒆𝒔 𝒂𝒏𝒚 𝒎𝒐𝒓𝒂𝒕𝒐𝒓𝒊𝒖𝒎 𝒕𝒉𝒂𝒕 𝒘𝒐𝒖𝒍𝒅 𝒑𝒓𝒆𝒗𝒆𝒏𝒕 𝒔𝒕𝒂𝒕𝒆𝒔 𝒇𝒓𝒐𝒎 𝒂𝒄𝒕𝒊𝒏𝒈... 𝑹𝒆𝒔𝒑𝒐𝒏𝒔𝒊𝒃𝒍𝒆 𝑨𝑰 𝒓𝒆𝒈𝒖𝒍𝒂𝒕𝒊𝒐𝒏 𝒔𝒉𝒐𝒖𝒍𝒅𝒏'𝒕 𝒑𝒆𝒏𝒂𝒍𝒊𝒛𝒆 𝒔𝒕𝒂𝒓𝒕𝒖𝒑𝒔 𝒇𝒐𝒓 𝒃𝒆𝒊𝒏𝒈 𝒂𝒄𝒄𝒐𝒖𝒏𝒕𝒂𝒃𝒍𝒆—𝒊𝒕 𝒔𝒉𝒐𝒖𝒍𝒅 𝒍𝒆𝒗𝒆𝒍 𝒕𝒉𝒆 𝒑𝒍𝒂𝒚𝒊𝒏𝒈 𝒇𝒊𝒆𝒍𝒅 𝒔𝒐 𝒓𝒆𝒔𝒑𝒐𝒏𝒔𝒊𝒃𝒍𝒆 𝒊𝒏𝒏𝒐𝒗𝒂𝒕𝒐𝒓𝒔 𝒄𝒂𝒏 𝒄𝒐𝒎𝒑𝒆𝒕𝒆 𝒇𝒂𝒊𝒓𝒍𝒚 𝒘𝒊𝒕𝒉 𝑩𝒊𝒈 𝑻𝒆𝒄𝒉." Thank you, Rep. Jamie Raskin, for ensuring the little guys have a voice in this debate. Here's what's at stake: ❌ The wrong path: A federal moratorium that freezes state action while Congress stalls—handing incumbents a regulatory moat. ✅ The right path: Let states like California set standards with laws like SB 53, creating momentum for a real federal framework built on accountability. This isn't a zero-sum game. Responsibility and innovation aren't enemies—they're partners. Without state leadership, we risk ceding AI innovation to the companies that can afford regulatory capture, not the ones building the future on safety, trust, and accountability. 🎥 Watch Rep. Raskin read our testimony

62

2 Comments

SEC Chair Paul Atkins just outlined the agency’s approach to artificial intelligence, and he is anchoring it firmly in a time-tested standard: materiality. By resisting the urge to create prescriptive "checklists" for AI disclosures, the SEC is correctly treating AI as a structural business development rather than an isolated novelty. My take: Statistical models do not possess an "intent" to break the law, which is why human judgment must remain at the center of enforcement and risk assessment. The mechanisms of market manipulation will undoubtedly evolve with GenAI, but our regulatory frameworks must remain technology-neutral. Here is why the SEC's principles-based approach hits exactly the right chord. https://lnkd.in/e6ktkzrs

14

1 Comment

For about 18 months, we've been evaluating a LLM AI system to help our team think more comprehensively and precisely about the programs and projects we design and execute for our family-run business clients. Today, our services touch projects and businesses in 6 states; all with different laws, rules, systems and approaches to human interaction. We thought it might be a tool that helps us be more intentional, systematic and cohesive. Because of concerns about fiduciary responsibility, confidentiality and potential negative impact, we approached this evaluation very deliberately with loads of guardrails. Of course, we follow the law and best practices (to the extent they yet exist), but we've also had to be really thoughtful about potential outcomes from what we "teach" it. We implemented a controlled experiment to evaluate both output and outcomes from our use of these tools. We record, measure and evaluate what's going in and coming out. Generally, we limit inputs to: hypothetical scenarios, anonymized deal structures and generic business models. We never use names or identifiable descriptors, precise financials, proprietary methods, real transactions or NDA-covered information. We're abstract and anonymous by default. We use the tools like a whiteboard, not a filing cabinet. We use them to "pressure-test" thinking, not to ingest source material. We've fully disclosed their use to clients and gave them an opt-out choice. We're not done evaluating, but we've already found some benefit in three ways: 1) extra caution about what gets used and not causes us to become more familiar with details and implications of our programs and projects, with obvious benefits to knowledge and understanding; 2) there's synergy with our institutional knowledge. We're generally not discovering new material, but we are being reminded "have you considered?" and 3) whether it’s a real outcome or a placebo effect, we're seeing team members work more closely as we make comparisons between with and without the tools. As a wrap-up, we asked for a summary of all our work in 2025 : A) Building Community Through Real Estate: You turned property development into purpose; structuring adaptive reuse, housing and accelerator projects that blend financial discipline with social outcomes for families and tenants. B) Designing Structures That Empower: You crafted a framework where families invest meaningfully in disability-support housing while preserving fairness, transparency and tax integrity. C) Turning Complex Ideas Into Working Models: You translated vision into structure- financial models, term sheets and operating agreements that made ambitious mixed use, housing and investor plans concrete and actionable.

3