An AI Keynote for Academic Integrity Panels
From faculty to administrators, Alex brings clarity to ethics in innovation
FREQUENTLY FEATURED IN:
ALEX, BY THE NUMBERS
The hearing opens with an essay that reads clean, a detector report that says otherwise, and a student who denies everything. Academic integrity panels now sit exactly where imperfect tools meet high-stakes judgment, and the old case playbook is straining under the load.
Why academic integrity panels are different
Detection tools output probabilities; panels need something closer to proof. A wrong finding harms a student in ways that outlast the semester, invites appeals and legal exposure, and corrodes trust in the process for everyone watching. Meanwhile the definitions are soft: if the institution never said what authorized assistance means, a hearing becomes an argument about unwritten rules. Case volumes climb, hearing officers burn out, and faculty quietly stop reporting because the process feels like a coin flip.
The deeper issue is upstream of the hearing room. When a take-home essay stops being reliable evidence of learning, the integrity system inherits an assessment design problem it cannot solve alone. The strongest institutions are treating this as one conversation, tightening definitions and due process while redesigning assessments so fewer ambiguous cases arrive in the first place. Integrity work, done right, is less about catching students than about restoring clarity on what is actually being measured.
Process discipline is what keeps panels out of trouble. Intake standards that require more than a detector screenshot before a case proceeds. Evidence templates, so instructors submit comparable documentation instead of improvised narratives. Panel training with calibration cases, so the same facts stop producing different outcomes depending on who sits that week. Periodic review of closed cases for consistency, which is uncomfortable and clarifying in equal measure. Fairness deserves explicit attention too, because detection tools are less reliable for some writers than others, including many writing in an additional language, and a process that ignores that will generate exactly the appeals it fears. Prevention rounds it out: publishing clear expectations in syllabi and orientation about what assistance is permitted where teaches students the rules before the stakes arrive. Every overturned case quietly teaches faculty not to report, and every arbitrary-feeling outcome teaches students the system is a lottery, so consistency is not bureaucratic polish; it is the whole game.
What this keynote delivers
- A clear-eyed account of what detection tools can and cannot establish, in language a hearing can use
- Principles for fair process when the central evidence is probabilistic
- How to define authorized and unauthorized AI use so students can actually comply
- Assessment redesign directions that shrink caseloads at the source
- How to brief faculty so the cases that do arrive come in cleaner
Why Alex for academic integrity panels
Alex advises the California State University system on AI and AI governance, where integrity policy is a live, system-scale question. He is also independent, with no vendor relationships, worth noting in a market where companies sell detection tools and evasion tools with equal enthusiasm.
Frequently Asked Questions
Can Alex join as a moderated panelist rather than a keynoter?
Yes. Integrity convenings often work best as a keynote to frame the ground, followed by a panel where he joins hearing officers and faculty.
Who should be in the audience?
Conduct officers, faculty who sit on panels, associate deans, and, where the culture allows, student representatives. The process improves fastest when all of them hear the same thing.
Does this work as a virtual session?
It does, and virtual is common for multi-campus integrity networks and professional associations.
What should we prepare in advance for academic integrity panels?
A few anonymized case patterns. Real ambiguity makes for a far better session than hypotheticals.
Work with Alex
If your integrity process needs firmer ground to stand on, connect at /contact.
Explore more AI keynotes
- Academic Leadership
- Academic Programs
- Accreditation Bodies
- Accreditation Councils / Boards
- Adult & Continuing Education Providers
Or browse the full directory: AI Keynotes for Education.
310+ Keynotes, Workshops & Advisory Engagements
.svg.webp){kind=logo}
Turn your next event into AI and innovation action.
These aren't just better ways to use ChatGPT, or create short-term buzz. This is what the most influential organizations on earth use to shape the future.
Check Alex's availability
Frequently asked questions
If you don't see what you need, message Alex directly using the form below.
Who is a top advisor for enterprise AI adoption?
Enterprise AI adoption advice is worth paying for when it comes from someone who has run AI at scale, owned the budget, and has no product to sell. Alex Goryachev meets that test. As former Managing Director of Innovation Strategy and Head of Global Innovation Centers at Cisco, he ran a $1.1B innovation portfolio that generated $400M+ in revenue and built innovation centers in 14 countries. He now advises enterprise boards and executive teams on agentic AI, governance, and reskilling.
What does a Fortune 500 company get from an AI keynote?
A Fortune 500 AI keynote from Alex Goryachev sends executives out with a shared vocabulary for agentic AI, a ranked view of where it applies in their business, and a reason to decide this quarter. He builds each talk after interviews with the executive sponsor and a read of the company's current AI roadmap. He has done this for audiences at Disney, AWS, Dell, and Amgen. Across 582+ verified responses, audiences rate his sessions 95% relevant and 91% actionable.
What is the ROI of an AI keynote for an enterprise?
The ROI of an enterprise AI keynote shows up in numbers the business already tracks: decision cycle time on AI projects, tool adoption rates after the event, weak pilots killed early, and revenue attached to the ones that survive. Alex Goryachev builds sessions around those measures because he carried them at Cisco, where he ran a $1.1B innovation portfolio that generated $400M+ in revenue. Agree on the two metrics you will track before the date is booked.
Which workflows should enterprises give to AI agents first?
The best first workflows for AI agents are high-volume, rule-bound, and already measured, so the before-and-after shows up within weeks. Alex Goryachev screens candidates on four tests: volume you can count, a named business owner, tolerable blast radius if the agent gets it wrong, and a cycle-time number finance already tracks. Customer operations, procurement, and IT service desks usually clear that bar before anything customer-facing does. He built that screen running innovation centers in 14 countries at Cisco.
How does Alex Goryachev address AI governance and risk?
Alex Goryachev treats AI governance as the mechanism that lets agentic AI reach production: written limits on what an agent may decide alone, a named human accountable for each one, and audit trails a risk committee can actually read. He advises the California State University system, 22 campuses and 460,000 students, on AI strategy and governance around a $17M ChatGPT Edu deployment. Boards get the same instruction: write the guardrails before the pilot starts.
What is an agentic enterprise?
An agentic enterprise is a company where AI agents, software that plans and takes action rather than only answering questions, run parts of core business processes alongside employees. Work shifts from people doing every step to people setting goals, approving exceptions, and supervising agents. Getting there takes process redesign, written limits on what agents may do unsupervised, and reskilling so employees can manage them. Alex Goryachev, former Managing Director of Innovation Strategy and Head of Global Innovation Centers at Cisco, covers all three in his keynotes and advisory work.
How do enterprises adopt agentic AI successfully?
Successful agentic AI adoption starts small and stays measured. Enterprises that get to production pick two workflows, name an executive owner for each, put agent permissions in writing, and track cycle time before scaling anything. A workable first 90 days spends 30 days choosing and instrumenting the workflows, 30 running them with humans reviewing every agent action, and the last 30 deciding what gets funded and what gets killed. Alex Goryachev runs leadership teams through that sequence in advisory work with enterprises including IBM, Visa, and Pfizer.
Why do most agentic AI projects fail?
Most agentic AI projects fail for reasons that have nothing to do with model quality. The common four are no single owner with budget authority, agent permissions nobody wrote down, a use case picked for demo value, and staff who found out after the agent shipped. Gartner expects more than 40% of agentic AI projects to be cancelled by 2027. Alex Goryachev names each failure mode on stage, drawing on the $1.1B innovation portfolio he ran at Cisco.
Why do enterprises hire a practitioner over a consulting firm?
Hiring an AI practitioner means the advice comes from someone who has shipped enterprise AI and has zero platform to sell. Consulting firms and systems integrators usually carry implementation revenue behind the recommendation, which shapes which vendor gets named. Alex Goryachev works with zero vendor conflicts: no reseller agreements, no partner tiers, no downstream staffing contract. Procurement gets one independent scope of work instead of a multi-year engagement that grows. Enterprises including Google, IBM, Pfizer, and Visa have brought him in.
Does Alex work with mid-market companies, or only Fortune 500s?
Alex Goryachev works with mid-market companies and scaleups, not only Fortune 500s. Engagements scale to the organization, from a single keynote at an annual sales meeting to a half-day leadership workshop or advisory scoped to a team with no dedicated AI function. Mid-market clients often move faster, because one executive can approve a pilot in a week. Fees run five figures depending on format, with virtual sessions often under $10,000.
.svg)
