The Mythos Dilemma
Source: Educating AI
Author: Nick Potkalitsky
Original source: https://nickpotkalitsky.substack.com/p/the-mythos-dilemma
Published: 2026-04-13
Source type: Essay
Private backup: the full article text is archived in the private repository at archives/articles/nickpotkalitsky-substack-com-the-mythos-dilemma.source.md. It is not published on the public Quartz site.
Summary
Nick Potkalitsky uses Anthropic’s leaked and later confirmed Claude Mythos Preview as a warning about school AI integration. Mythos is described as a highly capable cybersecurity model designed to find and exploit software vulnerabilities, and Potkalitsky argues that similar capabilities will eventually proliferate. The education concern is that districts are simultaneously being pushed toward more interconnected, API-dependent AI ecosystems involving LMSs, assessment tools, communication apps, student-information systems, and Google Workspace. The article does not argue against technology use; it argues that districts need harder questions about what they connect, why they connect it, and whether vendors have reckoned with the security cost of seamless integration.
Pull quotes
Seamless connectivity
The pitch was seamless connectivity. The promise was that more integration meant better outcomes.
— Nick Potkalitsky, The Mythos Dilemma
Integration as attack surface
Each connection is another surface. Each integration is another potential point of entry.
— Nick Potkalitsky, The Mythos Dilemma
The district question
That is not a reason to stop using technology. It is a reason to ask much harder questions about what they are connecting, why, and whether the vendor promising seamless integration has thought seriously about what seamless integration costs.
— Nick Potkalitsky, The Mythos Dilemma
Big ideas
- District AI work is a long-term redesign project
- AI tools should be judged by the work they will actually do
- Schools should start with learning values before choosing AI tools
Claims
- AI integration increases schools’ cybersecurity exposure
- Rushed school AI plans can worsen wellbeing and equity risks
- AI tool choices should be judged against stated learning values
Key evidence and examples
- Potkalitsky describes Mythos as an advanced model trained to find and exploit vulnerabilities, including zero-day flaws and autonomous end-to-end attacks on small-scale enterprise networks.
- He argues that Mythos-level capabilities will eventually proliferate beyond a small set of defensive partners.
- The article connects this capability trend to school infrastructure that increasingly links LMSs, assessment tools, communication platforms, student information systems, and productivity suites.
- Potkalitsky argues that districts should ask what they are connecting, why they are connecting it, and what seamless integration costs in security risk.
Education relevance
High relevance for district AI implementation, vendor review, student-data protection, cybersecurity risk, procurement, and avoiding naive “seamless integration” narratives.
Durability note
Durability: Medium. The Mythos-specific details are time-sensitive, but the broader warning that deeper AI integration expands school cybersecurity exposure is likely to remain relevant as district systems become more connected.