wa-law.org > bill > 2025-26 > SB 5956 > Original Bill
The legislature finds that an increasing number of public schools throughout the United States are beginning to use artificial intelligence, automated decision systems, and surveillance technologies in ways that can affect student discipline and school safety.
The legislature further finds that such tools can amplify existing disparities in student discipline and may result in disproportionate monitoring and punishment of Black students, indigenous students, students of color, students with disabilities, LGBTQ students, and other students protected under chapters 49.60, 28A.640, and 28A.642 RCW.
The legislature recognizes that Washington state has enacted student privacy protections, including in chapter 28A.604 RCW and RCW 28A.605.030, and that the office of the superintendent of public instruction has issued guidance on human-centered artificial intelligence in education. However, existing laws and guidance do not expressly address the use of artificial intelligence in automated decision systems in student discipline and surveillance.
Therefore, it is the intent of the legislature in this initial act to:
Prohibit the most harmful uses of artificial intelligence in automated decision systems and surveillance technologies in K–12 public schools, including predictive "risk scores" for individual students and real-time biometric tracking of students;
Ensure that automated decision systems do not replace the judgment of trained school personnel in discipline-related decisions; and
Direct the office of the superintendent of public instruction to update their guidance and the Washington state school directors' association to develop a model policy and procedures for school districts, charter schools, and state-tribal education compact schools consistent with these protections.
The definitions in this section apply to sections 3 through 6 of this act unless the context clearly requires otherwise.
"Artificial intelligence" means any machine-based system that, for explicit or implicit objectives, infers from the data it receives how to generate outputs, such as predictions, content, recommendations, or decisions, that can influence real or virtual environments.
"Automated decision system" means any computational process, including one derived from an artificial intelligence system, machine learning, statistics, or other data processing techniques, that makes or materially influences decisions or recommendations concerning a student.
"Biometric data" means data generated from the measurement or technical analysis of a student's physical, biological, or behavioral characteristics including, but not limited to, facial geometry, voiceprint, gait, fingerprints, or iris patterns.
"Imminent" means the state or condition of being likely to occur at any moment or near at hand, rather than distant or remote.
"Likelihood of serious physical harm" means a substantial risk that:
Physical harm will be inflicted by the student upon the student's own person, as evidenced by threats or attempts to commit suicide, or inflict physical harm on oneself; or
Physical harm will be inflicted by the student upon another, as evidenced by behavior that places another person or persons in reasonable fear of sustaining such physical harm.
"School service provider" has the same meaning as in RCW 28A.604.010.
"School surveillance technology" means any technology, service, or system used by or on behalf of a school district to monitor, track, or record student behavior, location, communications, biometrics, or online activity including, but not limited to, video analytics, network monitoring tools, and biometric systems.
"Student discipline-related decision" means any decision by a school district or its employees or contractors that results in or materially contributes to:
Suspension, expulsion, or emergency removal under chapter 28A.600 RCW;
Exclusion from class, activities, or transportation; or
Assignment to an alternative education setting for disciplinary reasons.
"Student personal information" has the same meaning as in RCW 28A.604.010.
An automated decision system may not be the sole or determinative basis for any student discipline-related decision.
No student may be emergency removed, suspended, expelled, referred to law enforcement, or assigned to an alternative education setting based solely on:
A prediction, score, or classification generated by an automated decision system; or
Data from a school surveillance technology, without independent human investigation and consideration of context.
Student discipline-related decisions remain subject to all requirements under state law, including protections against discrimination and requirements for due process.
A school district may not:
Use an automated decision system to generate a "risk score" or similar predictive classification for an individual student that purports to measure the student's likelihood of misconduct, gang affiliation, criminal behavior, targeted violence, or future disciplinary problems; or
Maintain internal lists or watchlists of students designated as likely perpetrators of violence or serious misconduct based in whole or in part on an automated decision system.
A school district may not enter into any contract with a vendor or school service provider that requires or authorizes the vendor or provider to engage in the activities prohibited under subsection (1) of this section.
School districts may not use biometric data to generate or infer emotional states, mental health conditions, sexual orientation, gender expression, gender identity, or other sensitive psychological or personal characteristics of a student.
The prohibitions in this section do not apply to:
The voluntary use of biometrics by an adult employee solely for secure access to facilities or devices; or
Uses required by federal law, provided that such uses are narrowly limited to what federal law requires and are not used for student discipline-related decisions.
Student personal information obtained or generated through an artificial intelligence system, automated decision system, or school surveillance technology may be disclosed to law enforcement only:
When required by state or federal law, including a court order; or
When there is an imminent likelihood of serious physical harm on school grounds or at a school-sponsored event, and the disclosure is limited to the minimum information reasonably necessary to respond to that threat.
Disclosures under this section must remain consistent with RCW 28A.600.475, 28A.605.030, and applicable federal law, including the federal family educational rights and privacy act of 1974, 20 U.S.C. Sec. 1232g.
Nothing in this chapter or RCW 43.386.080 may be construed to limit or impair the authority of a school district, school, or employee to take corrective action or impose student discipline as authorized under state law.
Nothing in this chapter or RCW 43.386.080 requires deletion of records that must be maintained or reported under state or federal law, including records required for student discipline data reporting and applicable records retention requirements under chapter 40.14 RCW.
Except as provided in subsection (8) of this section, a state or local government agency may not use a facial recognition service to engage in ongoing surveillance, conduct real-time or near real-time identification, or start persistent tracking unless:
A warrant is obtained authorizing the use of the service for those purposes;
Exigent circumstances exist; or
A court order is obtained authorizing the use of the service for the sole purpose of locating or identifying a missing person, or identifying a deceased person. A court may issue an ex parte order under this subsection (1)(c) if a law enforcement officer certifies and the court finds that the information likely to be obtained is relevant to locating or identifying a missing person, or identifying a deceased person.
A state or local government agency may not apply a facial recognition service to any individual based on their religious, political, or social views or activities, participation in a particular noncriminal organization or lawful event, or actual or perceived race, ethnicity, citizenship, place of origin, immigration status, age, disability, gender, gender identity, sexual orientation, or other characteristic protected by law. This subsection does not condone profiling including, but not limited to, predictive law enforcement tools.
A state or local government agency may not use a facial recognition service to create a record describing any individual's exercise of rights guaranteed by the First Amendment of the United States Constitution and by Article I, section 5 of the state Constitution.
A law enforcement agency that utilizes body worn camera recordings shall comply with the provisions of RCW 42.56.240(14).
A state or local law enforcement agency may not use the results of a facial recognition service as the sole basis to establish probable cause in a criminal investigation. The results of a facial recognition service may be used in conjunction with other information and evidence lawfully obtained by a law enforcement officer to establish probable cause in a criminal investigation.
A state or local law enforcement agency may not use a facial recognition service to identify an individual based on a sketch or other manually produced image.
A state or local law enforcement agency may not substantively manipulate an image for use in a facial recognition service in a manner not consistent with the facial recognition service provider's intended use and training.
A school district may not use a facial recognition service to engage in ongoing surveillance, conduct real-time or near real-time identification, or start persistent tracking of students.
During its regular review cycle, the office of the superintendent of public instruction shall, in consultation with students, families, educators, and communities disproportionately impacted by discipline and surveillance, update its guidance on human-centered artificial intelligence in K–12 education to reflect the prohibitions and requirements in chapter 28A.--- (the new chapter created in section 13 of this act) and to address the use of artificial intelligence systems, automated decision systems, and school surveillance technologies in connection with student discipline-related decisions and school safety.
By February 1, 2027, the Washington state school directors' association must develop a model policy and procedure that school districts, charter schools, and state-tribal education compact schools may adopt or adapt to implement the requirements in chapter 28A.--- (the new chapter created in section 13 of this act).
The Washington state school directors' association must maintain the model policy and procedure on its website at no cost to school districts.
The model policy and procedure must align with the office of the superintendent of public instruction's guidance described in section 9 of this act and must address:
Human oversight of artificial intelligence systems and automated decision systems;
Strategies to avoid discriminatory or disproportionately harmful impacts on students with protected characteristics under chapters 49.60, 28A.640, and 28A.642 RCW; and
Appropriate questions and criteria for evaluating vendors and tools that rely on artificial intelligence or automated decision systems in school settings.
Sections 2 through 7 of this act and RCW 43.386.080 govern school operation and management under RCW 28A.710.040 and apply to charter schools established under this chapter.
Sections 2 through 7 of this act and RCW 43.386.080 govern school operation and management under RCW 28A.715.020 and apply to state-tribal education compact schools subject to this chapter.