Effectiveness of an AI-Integrated Proactive Thought Control Intervention for Young Adults With Social Anxiety Disorder

The Objective is to Examine the Effectiveness of Ai Integrated Proactive Thought Control for Social Anxiety Disorder and Compare Its Effectiveness With Active Control Group and Cognitive Behavioral Therapy. Participants With SAD Will be Recruited From Educational Setting, Will Get 8 Therapy Sessions.

Participants with Social Anxiety Disorder will be recruited from educational settings . After screening and informed consent, participants will be randomly assigned to PTC , ACG or CBT .The therapy session will be 8 session once a week, and data will be collected through psychological questionnaires in two phases at start and end of intervention.

Study Overview

• Status: Completed
• Conditions: Social Anxiety Disorder (SAD)
• Intervention / Treatment: Other: AI-Proactive thought control; Other: Cognitive Behavior Therapy; Other: Active Control

• Study Type: Interventional
• Enrollment (Actual): 75
• Phase: Not Applicable

Contacts and Locations

Study Locations

• Pakistan
  • Punjab Province
    • Gujranwala, Punjab Province, Pakistan
      • Farah Shahzadi

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult
• Accepts Healthy Volunteers: No

Description

Inclusion Criteria:

• Were between 18 and 26 years of age, corresponding to the developmental period of emerging adulthood (Arnett, 2000).
• Met DSM-5 diagnostic criteria for Social Anxiety Disorder on the SCID-5-RV (First et al., 2015).
• Scored at or above the clinical cutoff of 30 on the LSAS-SR, indicating clinically significant social anxiety (Mennin et al., 2002).
• Were not receiving any other form of psychological intervention at the time of the study.
• Had sufficient English proficiency to understand instructions and complete the assessments (Flores, 2006).
• Provided written informed consent.
• Possessed basic computer literacy and typing skills required for the digital tasks.

Exclusion Criteria:

• Were younger than 18 or older than 26 years.
• Did not meet DSM-5 diagnostic criteria for Social Anxiety Disorder on the SCID-5-RV.
• Scored in the clinically significant range for general anxiety (BAI ≥ 16; Beck et al., 1988) or at or above the threshold for moderate depression (BDI-II ≥ 20; Beck et al., 1996).
• Were currently receiving psychological treatment for any psychiatric condition.
• Had a lifetime history of a severe psychiatric disorder other than Social Anxiety Disorder (e.g., psychosis, bipolar disorder).
• Had a medical or neurological condition that could interfere with task completion or valid responding.
• Were unable to complete the study procedures or assessments.
• Had no prior experience using a computer.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Treatment
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: Single

Number of Arms

3

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: proactive thought control; The PTC (Proactive Thought Control) arm an AI-integrated, computerized intervention that trains proactive thought control, the anticipatory maintenance of a positive goal representation used to override an automatically primed negative response before it is produced (Braver, 2012). Across eight sessions, each pairing a Free Association Task (single-word cues) with a Sentence Completion Task (unfinished social-evaluative stems).	Other: AI-Proactive thought control; The two paired tasks Free Association Task (FAT) and Sentence Completion Task (SCT) form the core of the AI-integrated Proactive Thought Control (PTC) program. FAT shows a social-threat word and the participant types the first associated word, training proactive control at the lexical level; SCT then gives unfinished social-evaluative stems to complete with a benign word, extending the override to the propositional level. Both run three blocks of thirty self-paced trials per session (720 over two weeks), so every threat trial forces a positive, cue-relevant representation that overrides a primed negative associate before advancing. AI integration relies on real-time validation: SiEBERT scores sentiment and MiniLM embedding confirm semantic relevance, enforcing a positive, cue-relevant construal each trial. Feedback reinforces this accepted responses earn points, rejected one prompt revision, and the threshold relaxes after four failures and pass after eight.
Active Comparator: Cognitive Behavior Therapy; The CBT group served as the gold-standard comparison condition, receiving an eight-session, individually delivered cognitive-behavioral protocol adapted from the empirically supported treatment of Hope, Heimberg, and Turk (2019). To fit the study's fixed eight-session frame, it followed the manual's accelerated outline, compressing the standard sixteen-to-twenty-session course while still covering most relevant workbook chapters in order.	Other: Cognitive Behavior Therapy; 8-session CBT protocol ( Hope, Heimberg & Turk, 2019 manual) The CBT arm received an eight-session protocol moving from psychoeducation and the three-component model of social anxiety with an individualized Fear and Avoidance Hierarchy (Sessions 1-2), through cognitive restructuring using disputing questions from automatic thought to rational response (Sessions 3-4), to repeated graded exposure using role-plays, distress ratings, and added cognitive tools, with the final session offering core-belief work and setting post-treatment maintenance goals (Sessions 5-8)
Active Comparator: Active Control Group; The Active Control group was a structurally and temporally matched, computerized active comparison condition that removed the intervention's active ingredient while preserving every non-specific feature. Participants completed the same two tasks (Free Association and Sentence Completion), the same cue/stem material, same three blocks of 30 trials per session (720 across eight sessions), same self-paced presentation, practice trials, on-screen appearance, and time on task, so the two conditions were equated on stimuli, dose, feedback salience, and engagement. The key difference was the omission of the proactive-thought-control contingency.	Other: Active Control; Participants did the same two tasks as the experimental group. The Free Association Task and the Sentence Completion Task , matched on stimuli, format, and time. The key difference was that the control removed the core training element: responses were not checked for meaning against the cue, and about half of the negative answers were accepted without correction, so the link between a threat cue and forcing a positive response was broken. In short, the control group did the same tasks and got the same exposure, but without the active thinking-training component.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Liebowitz Social Anxiety Scale-Self Report	Liebowitz Social Anxiety Scale-Self-Report; range 0-144; higher = worse (greater social anxiety severity)	Baseline (Day 1, before the first session) and Post-Intervention (within 7 days after the eighth session; the 8 sessions are delivered a weekly over a 8-week period
Core Belief Questionnaire (CBQ)	Core Belief Questionnaire-Self-Trait and -Other (17 items each, 6-point Likert); range 17-102 if items are scored 1-6; higher = more strongly endorsed negative core beliefs (worse)".	Baseline (Day 1, before the first session) and Post-Intervention (within 7 days after the eighth session; the 8 sessions are delivered a weekly over a 8-week period
Dot Probe task	milliseconds; more positive = stronger bias toward social threat	Baseline (Day 1, before the first session) and Post-Intervention (within 7 days after the eighth session; the 8 sessions are delivered a weekly over a 8-week period
word sentence association paradigm	higher = more threat-relevant interpretation bias	Baseline (Day 1, before the first session) and Post-Intervention (within 7 days after the eighth session; the 8 sessions are delivered a weekly over a 8-week period
Behavioral Avoidance Task	SUDS 0-100; higher = more subjective distress	Baseline (Day 1, before the first session) and Post-Intervention (within 7 days after the eighth session; the 8 sessions are delivered a weekly over a 8-week period

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
beck depression inventory II	range 0-63; higher = worse	Baseline (Day 1, before the first session) and Post-Intervention (within 7 days after the eighth session; the 8 sessions are delivered a weekly over a 8-week period

Collaborators and Investigators

• Sponsor: Farah Shahzadi

Study record dates

Study Major Dates

• Study Start (Actual): March 14, 2026
• Primary Completion (Actual): August 12, 2026
• Study Completion (Actual): August 14, 2026

Study Registration Dates

• First Submitted: August 29, 2026
• First Submitted That Met QC Criteria: September 3, 2026
• First Posted (Actual): September 9, 2026

Study Record Updates

• Last Update Posted (Actual): September 9, 2026
• Last Update Submitted That Met QC Criteria: September 3, 2026
• Last Verified: September 1, 2026

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Mental Disorders; Phobic Disorders; Anxiety Disorders; Phobia, Social; Behavior Therapy; Psychotherapy; Behavioral Disciplines and Activities; Cognitive Behavioral Therapy
• Other Study ID Numbers: GIFT/PSY/MS/2025/004

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: UNDECIDED

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No