Undue Influence in Litigation: Evidence, Mechanism, Causation, and Limits
Undue influence is often pleaded as though it were a conclusion. In serious litigation, it must be treated as a process.
An unexpected will, beneficiary designation, deed, gift, contract, power of attorney, fiduciary appointment, release, or settlement may justify close investigation. It does not establish how the result occurred. The litigation task is to determine whether evidence capable of proof connects the governing legal standard to a particular act, at a particular time, through a sufficiently supported mechanism—and to compare that explanation with the competing account of a voluntary decision under the applicable burden.
That inquiry requires legal and clinical discipline. The law defines the claim, allocates burdens, and determines which presumptions and remedies apply. Behavioral and medical evidence may help explain vulnerability, conduct, and decision-making function. It does not supply a substitute legal test.1
Start With the Governing Law, the Challenged Act, and the Operative Date
Before evaluating a diagnosis, relationship, or suspicious circumstance, counsel should identify:
the precise act or instrument being challenged;
the operative time as defined by governing law, including any legally relevant period in which the decision was formed—not merely the execution date when the law and facts make earlier conduct material;
the jurisdiction, cause of action, and controlling authorities;
the elements, burdens of production and persuasion, and standard of proof;
any statutory or common-law presumption and the facts necessary to trigger or rebut it; and
the remedy sought and the causal proposition that remedy requires.
The order matters. If the legal framework is left implicit, an evaluator can inadvertently convert age, dependence, cognitive impairment, active procurement, or an “undue influence” score into an additional element—or into a shortcut around an element that still must be proved.
Jurisdictional variation is substantial. For example, California’s elder-abuse definition directs attention to four nonexclusive groups of evidence: the alleged victim’s vulnerability, the influencer’s apparent authority, the influencer’s actions or tactics, and the equity of the result. It also states that an inequitable result, standing alone, is insufficient. Cal. Welf. & Inst. Code § 15610.70. California’s Probate Code incorporates that definition while providing that it supplements rather than supersedes common law. Cal. Prob. Code § 86.
Arizona uses a different statutory structure for governing instruments. Its statute begins with presumptions of capacity and freedom from undue influence, assigns the challenger an initial burden, and creates a rebuttable presumption of undue influence when specified relationship, procurement, and beneficiary facts are established. It also identifies proceedings to which specified subsections do not apply. Ariz. Rev. Stat. § 14-2712. These statutes illustrate why a national article can organize evidence but cannot announce a universal test. Rules may also differ for probate instruments, inter vivos transfers, multiple-party accounts, powers of attorney, contracts, and other challenged acts. Counsel must supply the law that controls the actual dispute. The linked statutory text was verified on August 10, 2026; counsel should confirm current statutes and controlling interpretations for the matter being litigated.
Build a Falsifiable Evidence-to-Act Chain
A useful theory of undue influence should identify each link in the proposed chain and the evidence that could weaken it. A theory that explains every possible fact—and cannot be affected by any contrary fact—is difficult to test or defend.
The sequence below is an analytic aid created to organize case evidence. It is not a diagnostic instrument, validated causal algorithm, score, legal test, or substitute for the elements and burdens supplied by counsel. Completing it establishes nothing. Its value lies in exposing missing links, contrary evidence, and assumptions that require proof.
1. Define the challenged act and alleged departure
State exactly what changed: ownership, control, beneficiary status, fiduciary authority, testamentary plan, litigation position, access to assets, or another legal relationship. Then identify the relevant comparison point. “Unusual” requires a supported baseline, not an assumption about what an older or ill person ordinarily would have done.
2. Reconstruct the person’s baseline and longitudinal course
The pertinent baseline may include prior estate plans, repeated expressions of intent, patterns of giving, conflict history, loyalties, risk tolerance, dependence, and decision-making habits. A single favorable or unfavorable encounter may not capture a fluctuating or evolving condition. Longitudinal evidence helps distinguish a stable value-based choice from a departure that arose during a period of vulnerability or control.
3. Specify susceptibility at the relevant time
Susceptibility is not proved by attaching a diagnostic label to the decision-maker. The question is functional and contextual: did a condition, loss, dependency, fear, cognitive limitation, psychiatric symptom, or social circumstance make this person susceptible in a functionally relevant way to the alleged tactic in connection with this decision?13 The companion article on susceptibility to undue influence develops that analysis in detail.
4. Identify the alleged influencer’s access, authority, and awareness evidence
Opportunity is relevant but not self-proving. Counsel should ask what access or authority the alleged influencer actually possessed; what evidence bears on awareness of the decision-maker’s vulnerabilities, assets, relationships, and prior plans; and whether the evidence shows use of that position. Where procurement is central, the separate analysis of suspicious circumstances and active procurement should control the deeper treatment.
5. Describe conduct, not character
The evidentiary focus should be on acts: filtering communications, controlling transportation or medication, restricting private access, creating urgency, supplying false information, threatening abandonment, arranging or directing professional meetings, scripting explanations, concealing the transaction, or rewarding compliance. A witness’s conclusion that someone was “controlling,” “predatory,” or “devoted” is less useful than a dated account of what the person said, did, knew, and prevented others from doing.
Litigators should distinguish emotional intensity from evidentiary value. Words such as betrayal, torment, fear, or devotion may describe a witness’s experience, but they do not establish the claim by themselves. They become useful when translated into conduct and placed in time.
The more difficult cases may involve false goodwill rather than overt threats. The alleged influencer appears protective, loyal, affectionate, indispensable, or professionally authoritative while controlling access, information, emotional interpretation, or the decision architecture. The decision-maker may defend that person and appear calm at execution. The relevant record may therefore extend well before the final ceremony.
6. Explain the causal mechanism
The central question is not whether vulnerability and suspicious conduct coexisted. It is how the alleged conduct operated on the identified vulnerability to produce the challenged act. The proposed mechanism should be specific enough to test against records and testimony. It may involve fear of losing care, distorted information, impaired ability to compare alternatives, dependency on a gatekeeper, repeated pressure during cognitive fluctuation, or another supported pathway. The mechanism must fit the operative time and the legal theory.
7. Test competing explanations
An adequate analysis considers explanations that do not require undue influence: a longstanding preference, estrangement, gratitude for care, tax or asset-planning advice, a reasoned response to family conflict, changed needs, ordinary persuasion, or independent professional advice. The existence of an alternative does not end the inquiry. Its fit with the complete record must be compared with the influence theory.
8. State the limits of the inference
Missing communications, unavailable witnesses, incomplete records, post-event accounts, uncertain timing, and biased sources may materially limit the opinion. Those limits belong in the analysis, not in a defensive paragraph added after a categorical conclusion.
Chronology Organizes and Tests the Causation Analysis
A chronology does more than arrange dates. It permits counsel to test whether the proposed cause preceded the result, whether the decision-maker’s functioning changed, and whether alternative explanations arose earlier or later.
A useful event chronology integrates at least five streams:
the person’s health, cognition, psychiatric state, and functional abilities;
relationships, caregiving, dependency, access, isolation, and conflict;
contemporaneous statements of intent and prior legal or financial acts;
the alleged influencer’s communications, conduct, and involvement; and
drafting, advice, execution, transfer, concealment, and subsequent use of the benefit.
The chronology should distinguish the date of execution from any legally relevant period in which the decision was formed. It should also identify missing evidence with a proper foundation: records ordinarily created in the relevant practice but not found, witnesses with material access who were not interviewed, unexplained communication gaps, and periods in which one person controlled the available information. Absence may limit an inference or identify discovery work; it does not prove concealment or misconduct by itself.
Separate Capacity, Susceptibility, Opportunity, Conduct, and Causation
These concepts may interact, but they are not interchangeable.
| Concept | What it can help explain | What it does not establish by itself |
|---|---|---|
| Decision-making capacity | Whether the person could perform functions required for the particular decision under the controlling standard | That another person exerted influence or caused the act |
| Susceptibility or vulnerability | Why a particular tactic might have had unusual force for this person at this time | That the tactic occurred, was legally undue, or caused the result |
| Opportunity or apparent authority | Whether the alleged influencer had access, leverage, or a position of trust | Use of that position or a causal effect |
| Conduct or procurement | What the alleged influencer did in relation to the decision or instrument | The decision-maker’s susceptibility or the legal consequence of the conduct |
| Causation | Whether the supported interaction between vulnerability and conduct explains the challenged act | The governing legal conclusion without application of the jurisdiction’s law |
A person may have capacity yet remain susceptible to pressure that exploits a specific dependency. A person may lack capacity without having been influenced by anyone. A beneficiary may participate in logistics without dictating the decision. Conversely, a facially orderly execution ceremony does not necessarily resolve what occurred during the weeks in which the decision was produced. Each proposition requires its own evidentiary support.12
Evaluate Evidence by Function, Timing, and Independence
Different sources answer different questions. Contemporaneous medical, legal, financial, and caregiving records may anchor dates, observations, access, and instructions, but they should be used only for what the writer observed and was positioned to assess. A routine “alert and oriented” notation does not resolve the disputed function; a private lawyer meeting may support process without disproving all earlier pressure.
Communications, drafts, metadata, call logs, calendars, account activity, transportation records, and access changes may test sequence, authorship, secrecy, and control. Witness accounts require the same provenance analysis: opportunity to observe, temporal proximity, consistency, interest, and distinction between direct observation and later narrative. Shared sourcing or extensive discussion may reduce apparent independence; it does not automatically invalidate the accounts.
Retrospective analysis can be appropriate when it identifies the operative question, reconstructs function at the relevant time, weighs source quality, and states uncertainty. It becomes vulnerable when it treats a later diagnosis as proof of an earlier condition, converts a cognitive-screening score into a legal conclusion, or selects only facts consistent with the retained side’s theory. The source discipline used in retrospective capacity assessment is relevant by analogy; it does not validate a separate susceptibility method.4 The companion susceptibility article addresses source quality and operative-time vulnerability in greater detail.
Use Behavioral Models as Maps, Not Verdicts
Behavioral models can improve issue-spotting and evidence organization. They can prompt counsel to examine susceptibility, authority, tactics, change over time, and outcome. They do not become governing law because their terminology resembles a statute, and they do not validate a conclusion merely because several models yield the same label.
That last point is especially important. Applying two frameworks to the same selected facts is not independent corroboration. The frameworks may share assumptions, domains, and source data. Any claimed “convergence” should therefore be described as organizational consistency—not as a measured probability, validated score, or substitute for external evidence.
The same caution applies to screening instruments. A tool can identify issues for further investigation but does not itself prove that influence occurred or caused a transaction. Published work developing an adult-protective-services screening tool expressly identified reliability and validity as subjects for future study.5 For a fuller discussion, see screening limits and susceptibility evidence.
Test Evidence That Supports and Weakens the Claim
High-quality analysis is symmetric. It asks what one would expect to observe if the influence theory were true and what one would expect if the act reflected an independent decision.
Consider a hypothetical record in which an older adult had long divided an estate equally between two children. After one child became the sole caregiver, that child restricted calls, attended all professional meetings, conveyed false claims that the sibling had stolen money, arranged a hurried new will, and received nearly the entire estate. Contemporaneous communications show fear of abandonment and repeated requests by the older adult to speak privately with the excluded child. That sequence could support a mechanism involving dependency, misinformation, isolation, and fear.
The same outcome would require a different assessment if the record showed that the older adult had independently described the reasons for the change months before the caregiving arrangement, initiated contact with counsel, met privately on several occasions, rejected proposed language, repeated stable reasons to disinterested people, and acted consistently with a documented history of conflict. None of those facts is automatically dispositive. Together they test the relative fit of the influence and voluntary-change explanations under the governing burden.
Evidence that may weaken an undue influence theory includes:
stable and repeated reasons expressed outside the alleged influencer’s presence;
meaningful private access to independent advisers and other relationships;
demonstrated ability to reject, modify, or delay the proposed act;
consistency with prior values, gifts, plans, or responses to documented conflict;
discrepancies between the alleged tactic and the decision-maker’s actual vulnerability;
a chronology showing the decision preceded the alleged conduct; and
direct evidence that materially contradicts the retained side’s key witnesses.
Evidence that may strengthen the theory includes:
a supported vulnerability that maps to the particular tactic;
evidence bearing on the alleged influencer’s awareness of and leverage over that vulnerability;
escalating control over information, access, care, or professional contacts;
false information, threats, concealment, coercive urgency, or repeated pressure;
a material departure arising during the period of control;
contemporaneous expressions of fear, confusion, ambivalence, or attempts to delay, refuse, verify, or contact others; and
post-transaction conduct only when independent evidence connects it to the earlier alleged control, rather than merely to receipt of a benefit.
These are evidentiary considerations, not a universal checklist. Their legal significance depends on the jurisdiction and claim. Top-level litigation presentation is sober. It does not rely on outrage. It shows the mechanism. That mechanism matters more than adjectives.
Define the Expert’s Assignment and Limits
An expert assignment should be framed narrowly enough to answer a real question. Depending on the jurisdiction, record, and permitted scope, a forensic psychiatrist may assist by:
reconstructing cognitive, psychiatric, medical, and functional conditions at the operative time;
explaining how an identified condition could affect retaining information, comparing alternatives, appreciating consequences, communicating a preference, acting despite dependency, or responding to a particular tactic;
evaluating the fit between the alleged mechanism and the longitudinal record;
identifying alternative explanations and evidence that would discriminate among them;
assessing the quality and limitations of the sources used; and
explaining why a diagnosis, screening score, or capacity opinion does—or does not—support a narrower functional inference.
General forensic-assessment guidance likewise emphasizes a defined referral question, appropriate collateral information, transparent reasoning, and disclosure of important limitations.6
The expert should not silently choose the governing legal test, treat a behavioral framework as law, or assume disputed conduct occurred. Whether an expert may use a legal term or address an ultimate issue depends on the forum, the governing authorities, the court’s rulings, and the precise question posed. In federal court, for example, Federal Rule of Evidence 702 requires the proponent to establish that expert testimony is helpful, sufficiently grounded, methodologically reliable, and reliably applied; Rule 704 provides that an opinion is not objectionable merely because it embraces an ultimate issue. Rule 704 does not make an unhelpful bare legal conclusion admissible or displace Rule 702 and the remaining evidence rules. Those federal rules should not be recast as a universal rule of state substantive law or expert scope.
A transparent opinion separates:
facts assumed or established;
source observations;
clinical or behavioral inferences;
the proposed causal mechanism;
reasonable alternatives;
degree of confidence; and
limitations that could change the opinion.
Questions Counsel Should Investigate Early
What exact act is challenged, and when was the decision actually formed?
Which legal elements, burdens, and presumptions control?
What was the person’s prior plan, and what admissible evidence establishes it?
What function or dependency allegedly created susceptibility?
What did the alleged influencer know about that vulnerability?
What specific words, acts, restrictions, misrepresentations, threats, or arrangements are alleged?
Which evidence connects that conduct to the challenged decision rather than merely to the relationship?
Who had private access to the person, and when?
What drafts, messages, metadata, account records, calendars, or access logs exist?
Which witnesses are independent, and which share information or interests?
What facts would be expected if the decision were voluntary?
What evidence materially contradicts the retained side’s theory?
What records or witnesses are missing, and how does their absence limit the inference?
Which opinions require expertise, and which are legal or factual determinations for the court or factfinder?
Need Case-Specific Analysis?
Early consultation can help counsel reconstruct the alleged mechanism of influence, test causation and competing explanations, and identify the evidence needed to connect vulnerability and conduct to the particular signature, transfer, appointment, release, or change in legal position. Learn more about forensic psychiatric analysis in undue influence matters.
Attorneys may submit a brief, non-confidential inquiry. Do not send records, privileged communications, protected information, or detailed case strategy before conflicts are cleared and an engagement is confirmed in writing.
This article is educational and is not legal or medical advice. Governing standards, burdens of proof, and evidentiary rules vary by jurisdiction, claim, and type of proceeding. Counsel should identify the controlling law; the court supplies the governing legal instructions, and the designated factfinder resolves the ultimate disputed issues.
Authorities and References
Legal authorities
Cal. Welf. & Inst. Code § 15610.70 (definition and nonexclusive evidence considerations; inequitable result alone insufficient), official text.
Cal. Prob. Code § 86 (Probate Code incorporation and preservation of common law), official text.
Ariz. Rev. Stat. § 14-2712 (burdens and statutory presumptions concerning governing instruments), official text.
Fed. R. Evid. 702, 704, in Federal Rules of Evidence (Dec. 1, 2025), official rules.
Clinical and professional literature
1. Herrmann N, Whaley KA, Herbert DJ, Shulman KI. Susceptibility to undue influence: the role of the medical expert in estate litigation. Can J Psychiatry. 2022;67(1):5-12. doi:10.1177/07067437211020616; PubMed.
2. Plotkin DA, Spar JE, Horwitz HL. Assessing undue influence. J Am Acad Psychiatry Law. 2016;44(3):344-351. PubMed; journal.
3. Peisah C, Finkel S, Shulman K, et al. The wills of older people: risk factors for undue influence. Int Psychogeriatr. 2009;21(1):7-15. doi:10.1017/S1041610208008120; publisher.
4. Shulman K, Herrmann N, Peglar H, Dochylo D, Burns C, Peisah C. The role of the medical expert in the retrospective assessment of testamentary capacity. Can J Psychiatry. 2021;66(3):255-261. doi:10.1177/0706743720915007; PubMed Central.
5. Quinn MJ, Nerenberg L, Navarro AE, Wilber KH. Developing an undue influence screening tool for adult protective services. J Elder Abuse Negl. 2017;29(2-3):157-185. doi:10.1080/08946566.2017.1314844; PubMed.
6. Glancy GD, Ash P, Bath EPJ, et al. AAPL practice guideline for the forensic assessment. J Am Acad Psychiatry Law. 2015;43(2 Suppl):S3-S53. PubMed; journal.