Professor Subodha Kumar warns AI-driven cyberattacks are an inflection point that may slow AI development
Summary
Host introduces Professor Subodha Kumar, who argues that recent outages and cyberattacks represent an inflection point driven by AI agents. Kumar predicts these incidents will accelerate into 2027, potentially shifting public sentiment to slow down AI development. He cites a Google report of an end-to-end attack by AI agents without human intervention. Kumar warns that systems are moving too fast and lack necessary protocols, increasing damage risks.
Episode sections
- 0:00 – 1:21
Introduction to AI-Driven Cyber Threats
The host introduces Professor Subodha Kumar, who argues that current outages and cyberattacks are an inflection point driven by AI agents. He predicts these incidents will accelerate into 2027, potentially shifting public sentiment to slow down AI development.
- 1:21 – 2:18
AI-Driven Attacks and System Vulnerabilities
Professor Kumar states that many recent cyber incidents involve AI implementations, citing a Google report of an end-to-end attack by AI agents without human intervention. He warns that systems are moving too fast and lack necessary protocols, increasing damage risks.
Claims worth verifying
Hackers carried out an end-to-end attack on Google without any human intervention, using AI agents.
Unidentified speaker (speaker_1) ·1:21
How to verify: Search for Google's official security blog or press releases from late 2026 (specifically around the podcast's release in Nov 2026) regarding AI-driven cyberattacks. Verify via Google's Cloud Security or Threat Intelligence reports. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
AI agents are becoming powerful enough to perform tasks independently without human intervention.
Unidentified speaker (speaker_1) ·1:40
How to verify: Review academic or industry reports on 'agentic AI' capabilities from late 2026. Check publications from the Information Systems Society or Temple University's Center for Business Analytics for technical assessments of autonomous AI task execution. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Professor Subodha Kumar is the founding director of the Center for Business Analytics and Disruptive Technologies at Temple University.
Unidentified speaker (speaker_0) ·0:40
How to verify: Visit Temple University's official website and navigate to the College of Liberal Arts or the specific Center for Business Analytics and Disruptive Technologies page to confirm his title and role. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Professor Subodha Kumar is an Information Systems Society Distinguished Fellow.
Unidentified speaker (speaker_0) ·0:40
How to verify: Check the Information Systems Society (ISSS) website's list of Distinguished Fellows for the relevant year (likely 2026 or earlier) to confirm his membership status. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Do your own research
- Verify the Google report cited by Professor Kumar regarding end-to-end AI attacks.
- Research historical trends in cyberattack frequency and attribution to AI agents.
- Review academic or industry publications on AI agent autonomy and security vulnerabilities.
- Check for regulatory responses or policy proposals aimed at slowing AI development due to security concerns.
- Analyze public sentiment data regarding AI safety and development pace.
- Identify specific protocols or standards currently lacking in AI system deployment.
Full transcriptexpand
- 0:00
- Unidentified speaker (UNKNOWN): Our
- 0:01
- Unidentified speaker (speaker_0): next guest makes the case that the current flurry of outages and cyber attacks happening right now is only the beginning of a significant inflection point and vulnerabilities that are being identified and exploited on a mass scale, primarily by AI. Specifically, hackers, both state and non-state actors, using
- 0:20
- Unidentified speaker (speaker_0): extremely powerful agenc agentic models to target everyone and everything. He also believes that we will likely see this accelerate from what is happening right now as we move into 2027. If that's the case, and if he's right about more and more critical sectors of the economy getting targeted and
- 0:40
- Unidentified speaker (speaker_0): disrupted, the most important takeaway from today's show is that this will dramatically shift the public conversation and outcry to slow down not just data center devices. development, but AI as a whole. Joining us on the show today is Sabodha Kumar. He's the founding director of the Center for Business
- 1:00
- Unidentified speaker (speaker_0): Analytics and Disruptive Technologies. He is also an informs Information Systems Society Distinguished Fellow and Professor at Temple University. What is your analysis on the surge that we've seen on outages, the cybersecurity-related incidences in recent
- 1:20
- Unidentified speaker (speaker_0): weeks?
- 1:21
- Unidentified speaker (speaker_1): They all lead to one very basic message. Many of these companies are dealing with AI implementations right now. Just this week, Google said that this hackers carried out the whole attack without any human intervention. This was an end-to-end attack by AI. Agents are becoming very powerful. And
- 1:40
- Unidentified speaker (speaker_1): more and more they are going to be independent in performing the task without human intervention. And we are very clear that it's going to be very powerful in a year or two. I think we are moving way too fast. The companies and the systems are not ready for some of these challenges. We have to follow certain protocols.
- 2:01
- Unidentified speaker (speaker_1): Otherwise, damages can be lot. bigger.
- 2:03
- Unidentified speaker (UNKNOWN): To
- 2:03
- Unidentified speaker (speaker_0): listen to this full interview and gain access to all of our premium content airing during the week, where we discuss markets and big investment trends, often well before they reach the mainstream news, go to financialsense.com and hit the subscribe button.
More from Financial Sense Newshour
IER: Don’t Blame Data Centers—War and Failed Policies Behind Rising Power Costs (Preview)
Episode discussion: Caleb Josso argues data centers are scapegoats for unreliable grid policies in blue states
Speakers’ claims: Host introduces post-midterm energy policy implications with guest Caleb Josso of the Institute for Energy Research. Josso claims blue states fail to innovate grid infrastructure by refusing reliable natural gas in favor of aggressive renewable mandates. He argues this creates fundamental reliability problems for the electricity grid. Josso asserts data center power demand is used as a scapegoat for these long-standing policy failures rather than being the core issue.
When Income Falls but Taxes Rise: The Widow's Penalty and the Case for Roth Conversions
Episode discussion: Crystal Colbert and Brendan McMurtrie discuss Roth conversions to mitigate the 'widow's penalty' and rising tax brackets.
Speakers’ claims: Crystal Colbert and Brendan McMurtrie analyze the tax implications of the 'widow's penalty,' where surviving spouses face higher tax rates due to narrower brackets and lower standard deductions. Colbert explains that larger traditional IRA balances lead to larger Required Minimum Distributions (RMDs) and higher Medicare IRMAA premiums. They present hypothetical scenarios showing how Roth conversions can reduce taxable income and preserve wealth for heirs. The speakers emphasize paying conversion taxes from outside the IRA and utilizing strategies like 'filling the bucket' to stay within lower tax brackets.