A client logs in from a hotel in another state. Another asks whether the session can continue while they drive. Your video platform suddenly offers an AI-generated summary. A documentation tool promises to turn a recording into a progress note in seconds. Later, a client sends a clinical concern through social media because they could not remember which portal to use.
None of these moments necessarily feels dramatic enough to trigger an “ethics emergency.” Yet each can change confidentiality, jurisdiction, informed consent, documentation, professional boundaries, or risk management.
That is why Digital Counseling has become a daily clinical competency rather than something therapists can address once in an annual policy review.
Telehealth and artificial intelligence can improve access, flexibility, and workflow. The problem is that convenience can make ethically important decisions feel routine. A clinician may experience the workflow as little more than clicking Start Meeting, allowing transcription, accepting an automated summary, and signing a note. Behind that seemingly simple process, sensitive clinical information may move through multiple systems and vendors.
The U.S. Department of Health and Human Services emphasizes privacy and security considerations when protected health information is used in telehealth. Federal telehealth guidance also highlights patient location, licensure, and emergency planning, while the American Counseling Association has published specific recommendations regarding AI, client welfare, transparency, professional competence, and human accountability.
The central question is therefore not:
“Is AI good or bad for therapy?”
A more useful clinical question is:
“What changes when this technology enters the therapeutic workflow—and who remains responsible when something goes wrong?”
Why Technology Creates a Different Kind of Clinical Risk
Traditional psychotherapy already requires clinicians to consider confidentiality, informed consent, boundaries, competence, documentation, cultural context, client welfare, and emergency response.
Technology does not replace those obligations.
It adds layers.
A virtual therapy session may involve:
- two different physical locations;
- multiple devices;
- broadband or cellular networks;
- a video platform;
- cloud infrastructure;
- an electronic health record;
- messaging systems;
- third-party integrations.
Add artificial intelligence, and the workflow may also involve:
- audio recording;
- automated transcription;
- data processing;
- generative summarization;
- model inference;
- data retention;
- subprocessors;
- automated recommendations.
Digital Counseling begins with understanding that workflow.
Therapists do not need to become software engineers, but they do need enough technological competence to ask clinically meaningful questions:
- Where is the client physically located?
- Am I authorized to provide treatment there?
- What happens if the client becomes acutely unsafe?
- Where does session information travel?
- Does a vendor create, receive, maintain, or transmit ePHI?
- Is a Business Associate Agreement required?
- Is audio recorded or temporarily processed?
- Is a transcript retained?
- Does AI generate clinical text?
- Does the client understand the technology being used?
- Can the client reasonably decline an optional AI feature?
- Who reviews errors before information becomes part of the record?
For HIPAA-regulated organizations, HHS explains that cloud providers that create, receive, maintain, or transmit ePHI on behalf of a regulated entity may be business associates, requiring appropriate contractual and privacy safeguards. HHS also emphasizes that a BAA does not eliminate the covered entity’s own risk-analysis and risk-management obligations.
Telehealth vs. AI: They Create Different Clinical Problems
Telehealth changes where and how treatment is delivered.
AI changes how information may be interpreted, generated, summarized, organized, or processed.
They overlap increasingly, but they should not be treated as the same technology.
| Clinical Issue | Telehealth | AI in Counseling | Clinician Responsibility |
| Primary purpose | Deliver care remotely | Assist with information, documentation, communication, or workflow | Understand what each technology actually does |
| Privacy risk | Video, messaging, environment, platform security | Transcription, processing, retention, integrations | Understand data flow |
| Consent | Explain risks and limits of remote care | Explain material AI use and its limitations | Keep consent meaningful |
| Clinical judgment | Some in-room information may be less accessible | Output may appear authoritative despite errors | Keep judgment human |
| Equity | Bandwidth, devices, private space, accessibility | Bias, representation, inaccessible interfaces | Assess impact on each client |
| Emergency response | Client may be geographically distant | AI cannot replace emergency evaluation | Maintain human crisis procedures |
| Documentation | Remote modality may affect documentation | AI may add, omit, or distort information | Verify the final record |
| Boundaries | Multiple communication channels | Automated responses may blur clinical availability | Clearly define the therapeutic frame |
A clinician can be highly competent at video therapy while remaining poorly prepared for an AI transcription tool.
Likewise, a practice can carefully vet an AI documentation system while failing basic telehealth procedures around client location and emergency response.
1. Your Client Is Suddenly in Another State
A client logs in and casually says:
“I’m visiting my sister this week, but we can still do our normal session.”
That sentence should trigger a location check.
Behavioral health licensure remains governed substantially at the state level, and Telehealth.HHS.gov explains that cross-state practice requirements vary. Possible pathways may include full licensure, temporary practice rules, reciprocity, interstate compacts, or telehealth registration depending on profession and jurisdiction.
This is where Digital Counseling becomes very practical.
A client’s permanent home address is not necessarily the only location that matters. Clinicians need procedures for determining where clients are physically located when jurisdiction is relevant.
A safer workflow
At the beginning of remote care—or according to applicable law, board rules, and practice policy—consider a brief verification:
“Before we begin, can you confirm where you’re joining from today?”
Then establish a clear procedure for situations in which a client is somewhere the clinician may not be authorized to practice.
Avoid universal statements such as:
“You can never see an established client when they cross state lines.”
The rules vary.
The appropriate question is:
What authorization applies in this client’s current jurisdiction?
2. A Crisis Happens Through a Screen
An office gives clinicians one major advantage in emergencies: they know where everyone is.
Teletherapy can remove that certainty.
A client could be:
- at home;
- in a hotel;
- in a vehicle;
- at work;
- on a college campus;
- staying with family;
- traveling.
The client then reports imminent risk.
The connection freezes.
Now what?
HHS guidance on telebehavioral health emergency planning recommends establishing an emergency plan and confirming information such as the patient’s location and relevant local emergency resources.
Your emergency procedure may need to include
- current client location;
- call-back number;
- local crisis resources;
- local emergency services;
- nearest emergency department when appropriate;
- an emergency contact where clinically appropriate and authorized;
- procedures for a disconnected high-risk session;
- documentation requirements;
- consultation/escalation procedures.
Example
A therapist is in Maryland.
The client’s chart lists Maryland.
Today, however, the client is visiting family in Georgia.
During session, the client reports escalating suicidal intent.
The clinician now faces both clinical risk and geographic logistics.
The time to design that process is not while the video screen is frozen.
3. “HIPAA-Compliant” Becomes the Entire Vendor Review
A platform homepage says:
SECURE. ENCRYPTED. HIPAA COMPLIANT.
That sounds reassuring.
It is not a complete risk assessment.
HHS explains that when a cloud provider creates, receives, maintains, or transmits ePHI on behalf of a HIPAA-covered entity or business associate, a compliant BAA may be required. HHS also explicitly states that OCR does not endorse, certify, or recommend particular cloud products.
Digital Counseling requires clinicians and organizations to look beyond the marketing language.
Before introducing a platform, AI scribe, transcription service, or clinical assistant, ask:
- What information enters the system?
- Is audio recorded?
- Is audio temporarily buffered?
- Is a transcript created?
- Is the transcript retained?
- Where is information stored?
- Who has access?
- Are subprocessors involved?
- Can information be used to improve models?
- Can data be deleted?
- What happens after account cancellation?
- Is a BAA offered when legally required?
- What happens during a security incident?
- Can necessary records be exported?
- What does the vendor actually promise contractually?
A BAA matters.
Encryption matters.
Access control matters.
But none substitutes for understanding the actual workflow.
4. Your Consent Form Says “Telehealth”—but You Just Added AI
Many practices created telehealth forms several years ago.
The form may discuss:
- internet disruption;
- privacy;
- technology limitations;
- emergency procedures.
Then the practice adds an AI note-taking tool.
Nothing changes in the consent process.
That is where trouble begins.
The American Counseling Association’s current AI recommendations emphasize transparency, competence, client welfare, limitations of AI, and maintaining accountability with qualified professionals. ACA’s recommendations also caution that AI-generated information can be inaccurate and that AI should not substitute for professional judgment.
Digital Counseling needs consent to function as an ongoing communication process, not merely an intake signature.
If a practice introduces:
- automated recording;
- transcription;
- AI summaries;
- ambient documentation;
- automated client messages;
- client-facing AI tools;
ask whether a reasonable client would consider that information important when deciding how to participate in care.
Better questions than “Did they sign the form?”
Ask:
- Does the client know what the tool does?
- Does it record anything?
- Does it generate a transcript?
- Is identifiable information being processed?
- How long does information remain with the vendor?
- Can the client decline?
- Is there a reasonable alternative?
- Who reviews automated output?
- How are inaccuracies corrected?
Clinical Events’ evergreen Informed Consent Guide for Therapists provides a broader framework for keeping consent active rather than treating it as paperwork completed once.
Important 2026 ethics-code note
The event’s learning objectives currently reference the 2014 ACA Code of Ethics, including Section H. ACA states that a revised final Code of Ethics is expected in Fall 2026. That makes ongoing verification especially important when publishing policies or training materials this year.
5. AI Turns a Complicated Session Into a Suspiciously Perfect Note
Here is where AI becomes very attractive.
Six emotionally demanding sessions are finished.
Six notes remain.
The AI-generated note is already organized:
Subjective. Objective. Assessment. Plan.
It sounds professional.
It is grammatically clean.
It takes 30 seconds to sign.
But a polished note is not necessarily an accurate note.
An automated system could:
- transform uncertainty into certainty;
- infer a diagnosis not made by the clinician;
- attribute a statement to the wrong person;
- omit important risk information;
- generate an intervention that never occurred;
- over-pathologize normal responses;
- misread culturally specific communication;
- add unsupported mental-status language;
- remove contextual details;
- create false precision.
Consider:
Client statement:
“I didn’t refill the medication because I couldn’t afford it.”
Automated interpretation:
“Client demonstrates medication noncompliance.”
The second sentence is shorter.
It is also clinically different.
The first describes an economic access barrier.
The second can imply behavior, attitude, or unwillingness.
The ACA’s AI guidance emphasizes that AI should remain adjunctive and that responsibility for decisions remains with licensed professionals. NIST’s AI Risk Management Framework likewise provides a framework for recognizing and managing risks associated with generative AI, including reliability, harmful bias, privacy, and human oversight.
A practical rule
Treat every AI-generated clinical note as an untrusted draft until a qualified clinician reviews it.
Clinical Events’ dedicated guide to AI Documentation Risks for Therapists explores these documentation-specific risks in greater depth.
6. The AI Output Is Plausible—but Still Wrong
Obvious AI mistakes are often easy to catch.
Plausible mistakes are harder.
Consider this client statement:
“I check the lock several times because my former partner has started appearing outside my apartment.”
An automated tool might summarize that as:
“Client demonstrates anxiety-related compulsive checking.”
That is possible.
It is not necessarily established.
Checking may also be connected to a genuine current safety concern.
Or suppose the note reads:
“Client dissociated while discussing childhood trauma.”
But what actually happened?
The client stopped speaking for 15 seconds and looked toward the floor.
Those are observable facts.
“Dissociated” is a clinical interpretation requiring additional assessment.
Keep these layers separate
| Documentation Layer | Example |
| Observation | Client paused and looked toward the floor |
| Client report | “I suddenly felt far away” |
| Assessment | Clinician assessed orientation and current distress |
| Hypothesis | Dissociative symptoms considered |
| Unsupported leap | Client dissociated because of trauma |
AI can help organize information.
It should not silently transform an observation into a diagnosis.
7. Bias Can Hide Inside “Efficiency”
Automation is sometimes experienced as neutral because a computer produced the output.
Automated does not mean neutral.
ACA’s AI recommendations specifically raise concerns involving diversity, equity, inclusion, inaccurate information, and uneven representation.
Potential clinical problems may include:
- culturally specific language being interpreted as pathology;
- racialized assumptions affecting risk descriptions;
- gendered assumptions entering summaries;
- disability-related communication being called “poor engagement”;
- nonstandard English being rewritten in a way that changes meaning;
- economic hardship being labeled “noncompliance”;
- spiritual beliefs losing cultural context;
- family structures being interpreted through narrow norms.
The clinician’s responsibility is not to prove that every algorithm is perfectly unbiased.
A more realistic responsibility is to identify where bias can enter and create safeguards:
- compare generated language with actual client statements;
- review labels;
- challenge unsupported certainty;
- examine differential impact;
- preserve cultural context;
- obtain consultation;
- keep diagnosis and treatment decisions human-led.
8. Technology Creates Access for One Client—and a Barrier for Another
Telehealth can remove transportation barriers and make remote interaction possible. HHS describes telehealth as allowing patients to interact with health professionals through video, phone, messaging, and related technologies.
But “available online” does not automatically mean accessible.
Consider clients who have:
- unstable broadband;
- limited cellular data;
- no private room;
- shared family devices;
- limited digital literacy;
- hearing or vision impairment;
- cognitive accessibility needs;
- language barriers;
- surveillance by an abusive partner;
- discomfort with recording;
- distrust of AI processing.
A client may technically have access to Zoom while having no safe place to talk.
Another may be able to attend video sessions easily but strongly object to an AI transcription tool.
Ethically responsible digital practice therefore asks:
What works for this client?
—not merely:
What is fastest for the practice?
The Clinical Edge Blog provides additional clinician-focused education across ethics, boundaries, documentation, informed consent, trauma, technology, and evidence-based care.
9. Your Screen Starts Competing With Your Client
Imagine the clinician’s visual field during a virtual session:
- client video;
- clinical record;
- chat window;
- risk alert;
- scheduling notification;
- AI-generated prompt;
- automated transcript;
- second monitor.
Every tool may be working exactly as designed.
The therapy can still become worse.
Digital Counseling should preserve therapeutic presence, not merely technical functionality.
A clinician might ask:
- Am I looking at the client or constantly at the EHR?
- Does the client understand why I keep looking away?
- Is an AI prompt narrowing my curiosity?
- Am I allowing automated suggestions to determine the direction of the session?
- Am I listening for meaning or waiting for the software to summarize it?
- Is technology making me more present—or merely faster?
Efficiency is valuable.
Presence is also part of quality care.
A workflow that saves 12 minutes of documentation but weakens the therapist’s listening may involve a tradeoff that deserves examination.
10. Your Clinical Boundaries Expand Across Five Different Apps
Teletherapy usually does not remain inside a video appointment.
Clients may communicate through:
- portals;
- email;
- SMS;
- scheduling apps;
- direct messages;
- social media;
- electronic forms;
- shared documents;
- client-facing AI products.
Each channel creates assumptions.
A client sends:
“I’m really struggling tonight.”
through Instagram at 11:47 p.m.
Did they believe it was monitored?
Did they believe it was confidential?
Was social media ever approved for clinical communication?
Will the message become part of the record?
What if the therapist does not see it until tomorrow afternoon?
Your digital communication policy should clarify
- approved channels;
- what each channel is used for;
- response times;
- emergency limitations;
- after-hours procedures;
- record-retention practices;
- social-media boundaries;
- technology-failure procedures;
- whether AI features are used.
For a broader clinical framework, Clinical Events has dedicated guides on Professional Boundaries for Therapists and Telehealth Ethics and Client Privacy.
What AI Can Assist With vs. What Should Stay Human-Led
AI does not have to be banned for clinicians to remain cautious.
The more practical question is where automation belongs.
| Task | Potential AI Assistance | Human Responsibility |
| Progress-note formatting | Organize clinician-approved information | Verify facts and clinical meaning |
| Administrative policy | Draft or simplify text | Verify accuracy and applicability |
| Appointment reminders | Automate routine communication | Set privacy and escalation rules |
| Psychoeducation | Create a draft | Verify evidence and tailor to client |
| Treatment-plan wording | Suggest language | Determine goals, diagnosis and client agreement |
| Risk assessment | Organize entered information | Conduct assessment and decide action |
| Diagnosis | Surface considerations | Licensed clinician completes diagnostic evaluation |
| Crisis response | Administrative support only | Qualified humans manage the crisis |
| Clinical recommendations | Generate possibilities | Clinician evaluates evidence and context |
| Final note | Flag missing information | Clinician owns the signed record |
ACA’s published AI guidance specifically recommends keeping licensed professionals accountable and cautions against relying on AI as a substitute for crisis response or independent mental-health diagnosis.
The Five-Layer Digital Risk Check
Before introducing a new feature or vendor, use a five-layer review.
1. Clinical Purpose
What problem are we actually solving?
Possible answers:
- improving access;
- reducing documentation burden;
- secure communication;
- accessibility;
- measurement;
- scheduling.
“Everyone else uses it” is not a clinical rationale.
2. Privacy
Ask:
- What data enters?
- Where does it go?
- Who receives it?
- Who stores it?
- How long?
- Can it be deleted?
- Who else has access?
For HIPAA-regulated organizations, assess whether a BAA is required—but do not stop with the contract.
HHS emphasizes both contractual safeguards and organizational risk analysis.
3. Consent
Ask:
Would a reasonable client care that this technology is involved?
If the answer is yes, determine:
- what should be explained;
- whether client choice exists;
- whether an alternative exists;
- how refusal affects treatment;
- whether consent should be updated.
4. Competence
Can the clinician identify when the technology is wrong?
If a therapist cannot evaluate an output, they cannot meaningfully supervise that output.
Competence means more than knowing how to click Generate.
It means understanding limitations.
5. Accountability
Ask one uncomfortable question:
Who is responsible when the output is wrong?
If the answer becomes:
“Well…the software wrote it,”
the workflow has a serious weakness.
This is the operational heart of Digital Counseling: technology may redistribute tasks, but it should not make professional accountability disappear.
A Day in the Life: How Small Technology Decisions Accumulate
Consider Maya, a fictional therapist.
8:55 a.m.
Her first client logs in from a vacation rental.
Maya notices the different background but does not confirm the client’s location.
10:00 a.m.
Another client begins discussing suicidal thoughts.
The video freezes.
Maya has the home address in the EHR—but does not know whether the client is actually home.
11:15 a.m.
The EHR announces:
NEW: AI NOTES — SAVE HOURS EVERY WEEK.
Maya clicks Enable.
She has not reviewed whether the tool records audio or retains transcription data.
1:00 p.m.
An automated note describes a client as:
“Noncompliant with medication.”
The client actually reported being unable to afford the refill.
2:30 p.m.
A client sends an Instagram message regarding a panic episode.
Maya responds because she wants to help.
The practice now has another de facto clinical communication channel.
4:00 p.m.
Maya wants help making an email sound more compassionate.
She copies the client’s message into a general-purpose AI tool without first considering whether identifiable clinical information is being transmitted to an inappropriate system.
No individual action necessarily felt enormous at the time.
The risk came from accumulation:
location + crisis logistics + vendor privacy + inaccurate documentation + boundary drift + uncontrolled data disclosure.
The solution is not fear.
It is a system.
Build the Policy Before You Need It
A digital-practice policy should be usable by therapists, supervisors, administrators, and clients.
Consider including:
- approved telehealth platforms;
- approved communication channels;
- location-verification procedures;
- cross-state practice workflow;
- emergency planning;
- privacy and security review;
- vendor approval;
- BAA procedures where required;
- approved and prohibited AI tools;
- transcription and recording rules;
- informed-consent procedures;
- AI-generated documentation review;
- data retention and deletion;
- accessibility accommodations;
- response-time expectations;
- incident reporting;
- staff training;
- periodic policy review.
Policies should also be revisited when technology, law, licensing rules, ethics codes, vendor features, or practice operations materially change.
That last point matters right now: ACA says its revised Code of Ethics is being finalized for publication in Fall 2026.
What Should Therapists Document in High-Risk Digital Situations?
Documentation should communicate clinical reasoning without becoming unnecessarily defensive.
Depending on the situation, clinically relevant documentation may include:
- service modality;
- client location when relevant;
- identity verification where required;
- consent discussion;
- significant technology problems;
- emergency procedures;
- risk assessment;
- consultation;
- client preferences;
- rationale for modifying a session;
- AI-output corrections;
- meaningful access barriers.
Avoid conclusions that exceed your expertise.
Rather than:
“Session complied with all applicable law.”
document the operational information and clinical reasoning that actually occurred.
Clinical Events’ guide to Clinical Documentation Mistakes Therapists Should Avoid offers a more detailed documentation framework.
When Should a Therapist Stop Using a Tool?
Technology discussions focus heavily on adoption:
Should we buy it?
Should we enable it?
Ethical practice also needs an exit strategy.
Consider pausing or discontinuing a tool when:
- vendor terms materially change;
- required safeguards disappear;
- important errors happen repeatedly;
- outputs cannot be independently verified;
- clients cannot meaningfully understand the workflow;
- accessibility problems worsen care;
- the tool interferes with therapeutic presence;
- data practices become unacceptable;
- the vendor handles security concerns poorly;
- the workflow exceeds clinician competence;
- safer alternatives become available.
NIST’s AI risk-management framework emphasizes ongoing governance, evaluation, and management of AI risk rather than assuming one approval decision permanently settles safety.
A 10-Question Ethics Check Before Clicking “Enable”
When a technology decision feels unclear, ask:
- What is the actual clinical purpose?
- What new risk does this create?
- What client information leaves our direct control?
- What would the client reasonably want to know?
- Can the client choose an alternative?
- Could this technology affect some populations differently?
- What happens when the output is wrong?
- Can a qualified person catch the error before harm occurs?
- What professional, organizational, licensing, payer, or legal requirement applies?
- Could I comfortably explain this workflow to the client, my supervisor, privacy officer, malpractice carrier, or licensing board?
The goal is not perfect certainty.
The goal is a defensible process.
Continuing Education: Digital Competence Is Becoming Clinical Competence
Clinical Events’ live virtual program Counseling in the Digital Age: Unveiling the Potential & Pitfalls of Telehealth and AI is scheduled for August 22, 2026, from 10:00 a.m.–1:15 p.m. Eastern Time and offers 3 Ethics CE Credits. The presenter is Marquita Maxwell, LPC, M.Div., NCC, ACS, BC-TMH. The program addresses telehealth evidence and risk, privacy, informed consent, AI in counseling, bias, professional identity, cultural responsiveness, digital disparities, and responsible technology integration.
For clinicians developing Digital Counseling competence, the value of training is not memorizing which platforms are acceptable today. It is learning a decision-making process that remains useful when tomorrow’s platform, documentation system, or AI feature looks completely different.
At the time of publication, clinicians can review the upcoming Counseling in the Digital Age training. Because individual training pages eventually become outdated, the evergreen Clinical Events schedule should also be used to find current CE programs, while the Clinical Edge Blog provides ongoing clinical resources.
Frequently Asked Questions
Is a secure video platform automatically HIPAA compliant?
Not necessarily in the sense clinicians sometimes assume. HIPAA compliance depends on the regulated entity’s complete workflow, vendor relationship, safeguards, contracts where required, risk analysis, access practices, and handling of protected information. HHS states that OCR does not certify or recommend specific cloud products.
Does having a BAA make an AI documentation tool safe?
No.
A BAA may be an important or required component when a vendor acts as a HIPAA business associate, but HHS also requires appropriate risk analysis and management by regulated entities.
Clinicians still need to understand:
data flow;
storage;
access;
retention;
deletion;
accuracy;
informed consent;
security procedures.
Can therapists use AI to write progress notes?
AI may assist with drafting and organization when its use is consistent with applicable privacy requirements, professional standards, employer policies, payer rules, client consent, and other relevant obligations.
The clinician should independently review the result before signing it.
ACA’s AI recommendations emphasize that responsibility remains with the licensed professional.
Can AI diagnose mental-health conditions?
ACA’s current AI recommendations do not support using AI as a replacement for qualified professional diagnosis. Diagnosis requires clinical judgment, context, cultural awareness, appropriate assessment, and human accountability.
Can therapists provide telehealth when a client travels?
Sometimes—but not automatically.
Cross-state requirements vary by jurisdiction and profession. Telehealth.HHS.gov identifies several possible pathways, including full licensure, temporary-practice provisions, reciprocity, compacts, and telehealth registration. Clinicians need to verify what applies to their specific situation.
What should happen if a telehealth client has a crisis?
Build the plan in advance.
HHS telebehavioral-health guidance emphasizes current location and local emergency planning so clinicians are not trying to reconstruct essential information in the middle of an acute crisis.
Should clients know when AI is used for their therapy notes?
When AI materially changes how sensitive clinical information is recorded, transmitted, processed, transcribed, summarized, or stored, transparency should be carefully considered as part of informed consent and ethical decision-making.
Exact legal requirements can vary by jurisdiction, profession, recording method, workflow, and setting, so clinicians shoul
Conclusion: Convenience Is Not the Same as Competence
Digital tools are no longer sitting outside mental-health practice.
Video sessions, client portals, cloud records, automated reminders, AI scribes, transcription tools, and generative systems are steadily becoming part of ordinary clinical workflows.
That normality is exactly why risk can become invisible.
The therapist who verifies where a client is located is not being unnecessarily bureaucratic.
The organization that investigates an AI vendor before uploading client information is not anti-technology.
The clinician who deletes half of an automated progress note and rewrites it is not defeating the purpose of AI.
They are keeping professional judgment inside the workflow.
Digital Counseling is safest when ethical safeguards become ordinary habits: verify location, prepare for emergencies, understand vendors, minimize unnecessary data, revisit consent, inspect AI output, monitor bias, maintain access, establish communication boundaries, and keep accountability human.
Technology can expand access.
It can reduce repetitive work.
It can support documentation.
It may even give clinicians back time that can be redirected toward clients.
But it should never make therapists less curious about risk, less attentive to context, less accountable for clinical decisions, or less present with the person on the other side of the screen.
The question is no longer:
“Should technology be part of counseling?”
It already is.
The better question is:
Can we use technology without allowing convenience to outrun clinical responsibility?
References / Authoritative Sources
For the WordPress reference section, I would use these core sources:
- U.S. Department of Health & Human Services (HHS) — HIPAA and Telehealth.
- Telehealth.HHS.gov — Licensing Across State Lines.
- Telehealth.HHS.gov — Creating an Emergency Plan for Telebehavioral Health.
- HHS Office for Civil Rights — HIPAA & Cloud Computing / Business Associate guidance.
- American Counseling Association — AI Work Group Recommendations for Counselors.
- American Counseling Association — ACA Code of Ethics revision status; final revised code currently expected Fall 2026.
- National Institute of Standards and Technology (NIST) — AI Risk Management Framework and Generative AI Profile.
