AI/Tech Talks at Thunderbird – September 2025 | Monthly Series

Date: September 25

Time: 1:30 – 2:30 PM

Location: Thunderbird Global Headquarters Building, Room 331

Tentative Agenda:

1:15 PM - 1:30 PM (15 min): Lunch – will be provided.

1:30PM - 2:30 PM (60 min): TBD.

Topic coming soon – but expect plenty of useful takeaways, fresh perspectives, meaningful connections, and yes – lunch and networking included!

-----------------------------------------------------------------------------------

Can’t join in person? RSVP to receive the Zoom link (ASU email required).

-----------------------------------------------------------------------------------

AI/Tech Talks Fall Update: Mixed AI Progress and Growing Regulatory Clarity

Dear Thunderbird and ASU Community,

Welcome back to the Fall semester! As we resume our AI/Tech Talks series, it's a perfect time to reflect on the significant AI developments from August that will shape our discussions this month.

Here are the key AI stories from this past month that deserve our attention:

GPT- Launch Meets Mixed Reception and User Backlash

OpenAI's much-anticipated GPT-5 launch on August 7 has received surprisingly mixed reviews from users and experts. While the company touted improvements in reasoning capabilities, many users have expressed disappointment, with thousands taking to Reddit to call the upgrade "horrible" and "a disaster." Critics like Gary Marcus noted that GPT-5 "still fails in all the same qualitative ways as its predecessors," while IEEE Spectrum highlighted the model's continued struggles with real-world tasks. The backlash was so significant that OpenAI had to reintroduce GPT-4o after over 3,000 petition signatures. This serves as a reminder that AI progress is not always linear, and hype often outpaces reality.

• Sources: Gary Marcus - GPT-5: Overdue, overhyped and underwhelming | IEEE Spectrum - GPT 5's Rocky Launch Highlights AI Disillusionment

Enterprise AI Reality Check: 95% of Corporate AI Pilots Are Failing

A sobering new MIT report revealed that 95% of generative AI pilots at companies are failing to move beyond the experimental stage. This finding challenges the widespread corporate enthusiasm for AI adoption and highlights the significant gap between AI potential and practical implementation. The report underscores the importance of realistic expectations and proper change management when integrating AI into business operations.

• Source: Fortune - MIT report: % of generative AI pilots at companies are failing

Global AI Regulation Gains Momentum with Clear Guidelines

August marked significant progress in AI governance worldwide. The EU AI Act's latest wave of obligations took effect, with member states required to designate national competent authorities by August 2. Meanwhile, the Trump administration's AI Action Plan emphasized "unbiased AI principles" and de minimis regulation to accelerate innovation. Several U.S. states, including California, Nevada, and Texas, enacted new AI laws addressing automated decision-making, deepfakes, and government AI use. This regulatory momentum signals growing global consensus on the need for clear AI oversight frameworks.

• Sources: DLA Piper - Latest wave of obligations under the EU AI Act take effect | Orrick - AI Law Center: August 2025 Updates

AI in Education: Bridging the Digital Divide

As schools nationwide grapple with AI integration, August highlighted both opportunities and challenges. New partnerships between educational technology companies and AI providers are bringing more AI tools into classrooms, while AI summer camps are working to close the growing AI literacy gap among students. However, concerns remain about equitable access and the need for proper teacher training to ensure purposeful tech integration rather than technology for technology's sake.

• Sources: NPR - Not all schools teach about AI. This camp wants to close the gap | EdTech Magazine - AI Training Options Open the Door to Purposeful Tech Integration

AI Security Concerns Rise as Misuse Detection Becomes Critical

Anthropic released new research on detecting and countering AI misuse, highlighting growing concerns about cybercriminals and malicious actors exploiting AI systems. This development underscores the critical importance of building robust safety and security measures into AI systems from the ground up, rather than as an afterthought.

• Source: Anthropic - Detecting and countering misuse of AI: August 2025

Historic Global Breakthrough: UN Establishes First-Ever AI Governance Framework

After six months of intense negotiations, all 193 UN Member States have unanimously agreed to establish two groundbreaking institutions: an Independent Scientific Panel on AI (40 experts providing evidence-based assessments) and a Global Dialogue on AI Governance (inclusive platform for international cooperation). Kenya's leadership representing 134 developing countries ensured Global South voices were central to this achievement. Building on September 2024's Global Digital Compact, this represents humanity's first universal agreement on international AI governance, with the first Global Dialogue scheduled for July 2026 in Geneva. As UN Secretary-General Guterres emphasized: "humanity must never be left to the 'black box' of an algorithm" – this framework signals a new era where AI governance is grounded in science, cooperation, and global solidarity rather than left to market forces alone.

Source: https://press.un.org/en/2025/sgsm22776.doc.htm

As we begin this new academic year, these developments remind us that the AI landscape is complex, nuanced, and far from the straightforward success story often portrayed in headlines. At Thunderbird, we remain committed to fostering critical thinking about AI's role in global business and society.

I was honored to participate in the Ai4 conference panel on "The Future of Education in the Age of AI" in August, which generated excellent discussions about balancing AI innovation with educational integrity and equity. These conversations will continue to inform our AI/Tech Talks series this year.

Looking forward to seeing you at our upcoming sessions as we navigate these important questions together.

Best regards,

Anjelina Belakovskaia

Associate Teaching Professor

Chair of the AI & Emerging Technologies Committee

Thunderbird School of Global Management

Arizona State University

Also check out other Workshops in Phoenix, Arts events in Phoenix, Nonprofit events in Phoenix.