Strategic Acquisition Brief · Confidential
MaxiFi is the deterministic engine that computes the provably correct lifetime financial plan — taxes, Social Security, longevity, lifetime consumption — the authoritative answer a probabilistic model cannot produce by sampling tokens. MaxiFi is the computed-correctness answer Perplexity can incorporate and stand behind — cited, reproduced, and defended — whichever model wins underneath it. Built over 30 years by BU economist Laurence Kotlikoff.
On April 9, 2026, Perplexity expanded its Plaid integration — positioned as an AI “personal CFO” — linking checking, savings, credit cards, and loans alongside brokerage to compute net worth, budgets, and debt-payoff across 12,000+ institutions, for every signed-in US/Canada user. That is a move into personal financial planning, the exact surface where a confident-but-wrong number does the most harm.
Perplexity’s entire franchise is accuracy and citations — the “trustworthy answer engine.” On the one question households care about most — how much can I safely spend, and how do I make it last? — an answer that is confidently wrong is worse than no answer. The open question is the entire pitch: can the answer engine whose promise is correctness give correct financial answers at consumer scale?
MaxiFi resolves it — the validated, deterministic engine that produces the mathematically correct lifetime plan, computed and not generated. It is the substance an answer engine cannot manufacture on its own, and the answer Perplexity can stand behind.
A large language model is a horizontal capability. In any function where a wrong answer is catastrophic, no serious operator ships the raw model to the user — a purpose-built application layer sits on top of it, encoding the domain’s rules and holding the model to them, turning raw generation into an action that is correct, defensible, and safe to deploy. It is already how high-stakes AI gets built: Intuit runs a deterministic tax engine under TurboTax’s AI and won’t let the model guess the numbers; patient-facing healthcare AI runs inside a safety layer, not on a raw model; and no one boards a plane flown by an unverified black box.
Personal financial planning is exactly such a function, and getting it wrong is its own kind of disaster: the retiree who runs out of money at 82, the family under-insured by a million dollars. It is also precisely where a model, left alone, fails — because it reaches for the same rules of thumb the incumbents use, and in this domain approximation is not “close enough”; it is wrong, in ways that compound every year to the household’s detriment.
Built from the ground up over thirty years, MaxiFi is the proprietary workflow that computes the correct, auditable answer under the actual tax and benefit rules — the function users actually want performed. The model doesn’t have it. The user doesn’t have it. The incumbents approximate it — and, in this domain, approximating it means getting it wrong, to the household’s cost. Value accrues to whoever owns that trusted workflow: the model layer commoditizes, while durable value moves up to the layer that owns the user’s trust and performs the function.
Perplexity’s promise is sourced, verifiable, cited answers. The public record on its accuracy is uneven. Even the best-performing answer engine in the Columbia Journalism Review (Tow Center) study — Perplexity — still returned incorrect citations roughly 37% of the time, against more than 60% across the field. The category leader on citations still fails on facts. A separate academic study (arXiv 2410.22349) found frequent hallucination and inaccurate citation — with overconfident phrasing the authors described as “hallucinations wrapped in the veneer of legitimate citations.”
There is a quieter point worth stating plainly. A citation tells a user where text came from; it does not make the lifetime financial math correct. And the legal direction is moving: courts are beginning to treat an AI engine’s synthesis as the engine’s own statement rather than a neutral pass-through of its sources. On a money question, then, the answer is the engine’s — not the citation’s. MaxiFi supplies what citation cannot: a computed, reproducible answer the engine can incorporate and stand behind, whichever model answers underneath.
At the sector level, the accuracy pressure on answer engines is widening: AI money advice offered without a fiduciary safeguard is beginning to draw its first suits across the industry. That is a category-wide dynamic, not a claim about any one company’s litigation history. The direction of travel is toward accountability for the substance of the answer, not the label on the product that delivered it.
Perplexity is building toward personalized money insight while disclaiming that it provides personalized financial advice. A read-only net-worth dashboard is one thing; a confident answer to “can I retire?” is another. The disclaimer does not travel with the screenshot a user takes of the number — or with the decision they make on it.
This is not a Perplexity-specific enforcement claim — it is the sector direction, and it does not depend on a fiduciary rule Perplexity sits outside of. A federal court has already treated an AI system as a product, not immune speech, allowing negligence and product-liability claims over its outputs to proceed. Personalized, account-linked money guidance sold for a subscription looks, in substance, like advice for compensation — a question the SEC can pursue on a simple-negligence standard, where registration is judged by what the tool does, not by what the disclaimer says. And once an industry is on notice that an output can be wrong at scale, the “novel technology” defense narrows for every participant in it — underwriters have already repriced it, with AI errors-and-omissions coverage that now specifically underwrites “hallucinations that cause financial harm.”
The threat is concrete and sober: wrong financial advice at consumer scale, after being on notice. At Perplexity’s reach, that is damages across the advised population, plus the reputational loss that compounds fastest for a company whose entire identity is correctness — “the AI you can’t trust with money.” The equity and franchise loss dwarfs the direct cost of any single wrong answer.
Most engines start from the aspirational question — “how much will you need?” — which manufactures a target number that is easy to state and hard to defend. MaxiFi alone starts from “what is the most I can spend with what I have?” — sustainable by construction. It is the question a household actually has, and the only framing that yields an answer an answer engine can stand behind.
MaxiFi is the financial-planning platform of Economic Security Planning, Inc., built over more than three decades by Professor Laurence Kotlikoff of Boston University. It uses consumption smoothing and dynamic programming to compute the single, mathematically optimal lifetime plan — solving simultaneously across Social Security strategy, Roth-conversion sequencing, withdrawal order, and the full tax code.
Goals-based tools answer “What is the chance you hit your number?” MaxiFi answers “What is the optimal path, and how much can I spend today without jeopardizing tomorrow?” It is not a better simulator. It is a different class of engine — the deterministic, computed answer Perplexity can incorporate and cite.
Prof. Laurence Kotlikoff — William Fairfield Warren Professor at Boston University; Harvard Ph.D.; former Senior Economist on the President’s Council of Economic Advisers; named by The Economist among the 25 most influential economists. Larry Kotlikoff intends to stay on with the acquirer in whatever capacity best serves the product — architect, spokesperson, advisor.
Taught by Nobel Laureate Robert Merton at MIT Sloan as an “outstanding science-based lifecycle and retirement management platform” — Merton does not endorse products; he teaches MaxiFi as the reference engine. Featured in Bankrate’s “Best financial planning software of 2025” roundup (May 5, 2025), cited as best for near- and long-term tax planning and the decumulation phase.
Patent-winning algorithms refined over 30+ years, built from economic theory rather than scraped text — exactly the kind of intellectual property a probabilistic model cannot reverse-engineer by sampling tokens.
MaxiFi is already in the wealth-advisor channel via its Pro subscription — fee-only planners use it as top-of-funnel lure and client-retention glue. The acquirer inherits a paying installed base that extends naturally to Perplexity’s consumer reach.
Perplexity is not a model trainer — it is the distribution and answer layer. It is explicitly model-agnostic, sitting downstream of both Anthropic Claude and OpenAI GPT as well as its own Sonar models. That is exactly why the fit is strong: MaxiFi incorporated into Perplexity gives it one authoritative, computed financial answer regardless of which model wins underneath — a correctness core that does not depend on which lab is ahead this quarter.
A money question is a different kind of intent than the rest of what Perplexity answers: it needs a computed answer, not a sampled one. Incorporated into Perplexity, MaxiFi supplies that answer — one it can cite, reproduce, and defend, however the underlying integration is engineered.
Perplexity sits downstream of both Anthropic and OpenAI. Incorporating the correct-answer engine is the same bet regardless of which lab wins the model layer underneath — and it is a capability neither of its model suppliers can revoke. It is the one finance capability that stays durable across every shift in the frontier.
The result is cybersecurity-style assurance: Perplexity can tell its users, its publishers, and its investors that money answers are computed, verifiable, and reproducible — the same way a security vendor lets its customers stand behind what it ships.
The neutral national-press datapoint is the cleanest: CBS MoneyWatch (May 7, 2026) ran an identical retirement prompt — a 50-year-old single woman retiring at 65 — through Claude, ChatGPT, and Perplexity. The verdicts diverged and Perplexity was the most pessimistic. Larry’s Economics Matters Substack — 137,000+ subscribers — has run a six-post sequence testing named engines against MaxiFi on dollar-specific household problems. The variance across engines on one identical prompt is the proof.
“Asked whether a 50-year-old single woman could retire at 65, Claude, ChatGPT, and Perplexity diverged — Perplexity the most pessimistic. Kotlikoff: AI ‘may do more harm than good,’ mishandling Social Security and wrongly averaging longevity instead of using maximum life expectancy.”
The divergent-verdict story →“The AI said John and Jane can spend approximately $52,000 per year in discretionary spending. MaxiFi’s demonstrably correct answer — verifiable by inspecting its reports — is $63,382.”
Read the head-to-head →“AI’s best hope of providing accurate economics-based planning is by serving as a front end guiding data entry and using MaxiFi as the back end to produce precisely correct, not clearly pretend, results.”
Read the structural argument →“The median household leaves $182,370 of lifetime Social Security on the table. AI tells Jane a job change adds at most $35K in lifetime benefits when the right answer is $168K.”
Read the Social Security test →Acquiring MaxiFi acquires the megaphone these pieces ship from — pointed, with credibility no one in the category can match, at the consumer-finance surface Perplexity just opened. Larry intends to keep contributing to the product and to stay on as spokesperson, turning a category critic into Perplexity’s correctness narrator. The CBS divergent-verdict finding is the named, neutral proof; the Substack series is the dated, dollar-specific record behind it.
The same mechanism — being right about a user’s money — is revenue today and a durable, higher-quality earnings stream tomorrow. It runs as one funnel: attractant → adhesive → loyalty → continuity. Correctness draws the user in on the highest-stakes question they will ever ask an answer engine; correctness keeps them, because a wrong answer on money is disqualifying and a right one is not replaceable with a cheaper competitor; retention becomes recurring, lifelong engagement; and that same engagement is what a market pays a premium multiple for, because it is durable and hard to copy.
Questions about one’s own money are the highest-intent, highest-trust, highest-willingness-to-pay questions a user ever asks an answer engine. Being right is the attractant, the adhesive, and the source of loyalty — recurring, lifelong engagement, and qualified demand for everything else Perplexity offers.
A correct-by-construction engine retires the largest overhang on the money-answer story: being confidently wrong with people’s money, at scale, at the exact moment AI money advice with no fiduciary safeguard is drawing its first suits, sector-wide. And there is exactly one MaxiFi — in a rival’s hands, it strengthens their answer and weakens Perplexity’s.
The market pays a higher multiple for a dollar of earnings that is defensible and low-risk. MaxiFi makes the same revenue from Prong 1 proprietary, un-copyable, and trust-based — and removes a tail risk in the same motion. A moat plus a removed risk is what re-rates a multiple.
The attractant→adhesive→loyalty→continuity funnel is the same mechanism twice: it is the revenue engine today, and it is the durable, high-quality earnings stream the market re-rates tomorrow. One asset, financing both sides of the story.
Perplexity sits downstream of both Anthropic and OpenAI. Incorporating the correct-answer engine is the same bet regardless of which lab wins the model layer — and it is a capability neither model supplier can revoke. A 30-minute briefing makes the case concrete: a live demonstration where MaxiFi solves a household’s lifetime plan while the leading models are asked to match it. The gap is the entire thesis.
MaxiFi is being offered through a focused strategic process. The preference is an acquisition — that is where the strategic value sits. Continuity de-risks it: Larry Kotlikoff intends to stay on with the acquirer in whatever capacity best serves the product — architect, spokesperson, advisor. The path is short, and the strategic payoff — product integrity, a model-agnostic moat, and competitive denial — is immediate.