Outreach Templates_

Hi [Name],

I noticed [Company] offers [therapy/coaching/wellbeing features]. Quick question for you:

When a regulator or auditor asks how your AI captured emotional context from a user session — and what decisions were made based on it — what documentation do you provide?

The EU AI Act has begun applying prohibited-practice restrictions (from Feb 2, 2025), and teams are tightening how they document sensitive human context. I'm seeing products that handle sensitive human context start to think about this differently. Not because they're in violation, but because demonstrating the distinction between interpretation (structuring what was emotionally significant — expressed or interpreted) and inference (deriving hidden states from behavioral signals) is increasingly showing up in enterprise due diligence.

Would be curious how [Company] is approaching this. Happy to share what we're seeing in the market if useful.

Best,
[Your name]

Hi [Name],

I work with compliance teams at companies that handle emotional context in their AI systems — therapy platforms, wellbeing apps, coaching tools.

One question I'm hearing from risk officers: if a user disputes how their emotional state was characterized in your system, can you produce audit-friendly, tamper-evident documentation showing:
- What explicit expression was the basis for that characterization
- What model and version performed the interpretation
- What governance metadata was attached
- Whether the user can verify the interpretation against their original words

Many teams can reconstruct parts of this from logs; fewer can produce it as a portable, verifiable record.

Is this something on [Company]'s radar? Would be happy to share what frameworks others are implementing.

Best,
[Your name]

Hi [Name],

Quick question on [Company]'s emotional AI compliance approach:

Article 5 includes restrictions on certain forms of emotion recognition/inference in specific contexts (including workplace and education). The exact scope — particularly around what constitutes "inference" vs. structuring expressed or interpreted content — is an area many legal teams are still evaluating.

I work with platforms that handle emotional context, and one thing I'm seeing: companies that can demonstrate interpretation-based architectures (structuring what users explicitly communicate) may be viewed differently in audit and regulatory review than those using inference (deriving states from involuntary signals).

Not legal advice — we're focused on producing better records and clearer boundaries.

Is this a distinction [Company] has mapped internally? Happy to share a brief on how other platforms are documenting this if helpful.

Best,
[Your name]

Hi [Name],

Following up briefly — I put together a short brief on the interpretation vs. inference distinction for emotional AI systems. It covers:

- What a regulator/auditor typically asks for
- What to store as a governed receipt (expressed → interpreted → policy)
- How to make it portable + tamper-evident

Would you find this useful? Happy to send if so.

Best,
[Your name]