As a follow up to our CIO Forum event on Thursday, June 20th, 2024, we would like to give special thanks to our speaker, Paolo Pelizzoli, CTO of Platform Engineering at The Clearing House, and President of Alakrity LLC, for his innovative discussion on the state of AI today.
A few takeaways:
Paolo led a discussion on the potential and challenges of artificial intelligence (AI), with a focus on data privacy, security, and its integration into various sectors. Paolo presented on the growing importance of AI in managing large data sets, its potential to revolutionize systems, and the need for proper data strategies and security measures. The group also discussed the potential risks of AI, its impact on the job market, and the evolving role of the chief data officer in managing its implications.
Recommendations:
- Get going: Organizations need to get staff familiar in a structured managed fashion, but with some level of controls and boundaries.
- Organizations need a Data Strategy as well as an AI Strategy as data of course feeds the AI models and both require oversight and management to be effective.
* Assign an AI expert to monitor the use of AI within the organization.
* Provide training on how to use AI safely and effectively.
* Establish guidelines for when AI can be used to modify transactions.
* Develop a plan for protecting sensitive information from being incorporated into AI models.
Discussion items:
- Can AI and data privacy co-exist?
- Can your data be protected from ML (machine learning) and AI?
- There are many models and AIs surfacing, as these new capabilities rapidly evolve.
- (Large Language Model) requires extensive training on massive textual data sets.
- Pressure to ingest data, risks of including PII, our habits can be incorporated into these AI models, how can we block inappropriate data ingestion, data insights and surfacing of inappropriate insights. How do we restrict and manage these models, in a financially feasible manner? My Data?
- There is a blurring of lines. We’re seeing deep fakes, voice-overs, targeted misinformation, crossing lines of Intellectual Property protection, where IP leakage gets assimilated into AI models. The legal system is playing catchup with the fast pace of AI. There are burgeoning capabilities for flagging data explicitly to be excluded, with the legal system stepping into formalizing and starting to enforce. Other emerging capabilities include attestation, and chip-level encryption.
- 3rd party risks
- Is the AI and underlying data used ethically? Does it travel overseas? Can I export my data, or have it removed or redacted? Would that then undo the weightings that were affected by my data? Can I trust these capabilities? What happens during M&A? How do we manage and control risk of bad actors?
- AI is surfacing across a wide range of applications and affect business processes.
- How do we debug; how do we manage AI code and how do we manage these ecosystems?
- Speed: We are clearly within a hype cycle, but it’s real and evolving fast, and actually accelerating.
- Controls: The controls are being developed, as are the legal controls.
- Inherent Risks: The risks are real: ethical, personal, organizational, governmental and pan-civilizational.
It was great seeing all of you who joined us! Have a wonderful summer!
-malka
Malka Treuhaft
Executive Director East Coast CIO Forum &
President
Truision Inc.
646.942.2625 (office)
917.589.1069 (mobile)
718.375.1529 (fax)
www.truision.com
