KOHEN AVOCATS FIRM · PARIS 16
India’s Extradition Treaty Expansion: Why a Signature Does Not Guarantee Surrender

Hassan KOHEN acts in urgent matters, from the police station to the assize court. Initial strategic analysis: €80 incl. VAT, personal reply within 24 hours.

100% confidential · Legal professional privilege · No obligation

Article generated by artificial intelligence, through a process designed and supervised by the firm

Source: French Court of Cassation – Judilibre and Légifrance open data

Paris Bar Police custody, investigation, assize court CNB listing avocat.fr
Hassan Kohen, criminal defence lawyer in Paris
Hassan KOHEN
Attorney at the Paris Bar

India’s Extradition Treaty Expansion: Why a Signature Does Not Guarantee Surrender

Speak to a lawyer 06 89 11 34 45

India is widening the map of its extradition cooperation. On 26 August 2026, Hindustan Times reported that the Union Cabinet had approved the signing of extradition treaties with Kenya, Kazakhstan and the Kyrgyz Republic, while also giving ex-post-facto approval to the treaty signed with Seychelles during Prime Minister Narendra Modi’s June visit. The same report put the projected treaty network at 52 countries and drew a necessary distinction between formal extradition and the wider number of fugitives returned through deportation or other forms of cooperation.

That distinction is the legal story. A Cabinet approval can open a diplomatic channel, but it does not itself produce an enforceable surrender. The text of each treaty, the constitutional steps needed for entry into force, the requested State’s arrest procedure, dual-criminality rules, evidence requirements, political-offence safeguards, human-rights controls and the final decision on surrender remain decisive. A person located in France would not be handed over merely because India has signed a new instrument. French courts and the French executive would still apply the applicable treaty and the Code of Criminal Procedure.

For an international audience, the new network therefore raises a practical question: what changes when a State moves from informal returns and diplomatic requests to a treaty-based extradition process, and what can still stop surrender? French and European procedure provides a useful answer. It shows why a treaty increases the legal tools available to prosecutors while simultaneously creating a precise framework in which the requested person can challenge identity, classification, documents, human-rights risks, the sentence sought and the conditions attached to any eventual transfer.

I. What do India’s new extradition treaties actually change for fugitive transfers?

A. Does a Cabinet approval make an extradition treaty operational?

The first mistake would be to treat a political announcement as a completed surrender mechanism. The reported Cabinet decision concerns the next stage of cooperation: signing and, according to the report, domestic approval of instruments concerning Kenya, Kazakhstan and the Kyrgyz Republic. It also concerns approval after the fact of the India–Seychelles treaty signed in June. That is important, but it does not answer the questions that determine whether a request can be executed tomorrow. The dates of signature, ratification, notification and entry into force must be identified for each bilateral instrument. So must the designated central authorities, the required documents, the language rules, the time limits for provisional arrest and the grounds on which the requested State may refuse surrender.

India’s own treaty-listing practice is the right starting point for verification. The Ministry of External Affairs list of extradition treaties and arrangements is a useful map, but a list is not a substitute for the operative treaty text. A bilateral instrument can be signed while implementation remains dependent on ratification, exchange of notifications or domestic publication. The official Indian material on the June visit to Seychelles confirms that an extradition treaty was among the instruments exchanged. That confirms the diplomatic event; it does not, without the treaty’s entry-into-force clause and the necessary notifications, establish that every future request will be accepted.

The public record concerning Kazakhstan illustrates the same point from the other side. A presidential decree published in Kazakhstan in February 2026 approved a draft India–Kazakhstan extradition treaty. The draft’s opening obligation is expressly subject to the treaty’s conditions and the national law of the contracting States. The legal effect is therefore layered: the States agree to build a surrender channel, but a requested person remains protected by the procedural law of the State where that person is found. This is precisely why the word “treaty” should not be confused with the phrase “automatic transfer”.

The France–India convention published in the French Official Journal offers a concrete illustration of the sequence. Its Article 24 requires each State to notify the other that its constitutional procedures for entry into force have been completed. It then provides that the convention enters into force on the first day of the second month after the last notification. The operative language is clear: “Chacun des deux Etats notifiera à l’autre l’accomplissement des procédures requises par sa Constitution pour l’entrée en vigueur”. That clause is not decorative. It separates a signature ceremony from the moment when an authority can rely on the convention as the legal basis for a particular request.

The same convention shows how a treaty turns political cooperation into a document-driven process. Article 1 states that the contracting States undertake to surrender persons prosecuted for an offence or sought to serve a prison sentence, but only “selon les règles et dans les conditions déterminées dans la présente convention”. Article 9, read with the provisional-arrest provisions, determines what must accompany a request. Article 21 then provides that, unless the convention says otherwise, the law of the requested State governs the extradition procedure and provisional arrest. That combination is the operational core of a treaty: an international obligation, bounded by a negotiated list of offences and safeguards, executed through the requested country’s own courts and officials.

The reported Indian expansion should consequently be read as an infrastructure decision. It can make future requests more predictable than a purely diplomatic approach. Prosecutors can point to a defined legal basis instead of asking the other State to act through an informal arrangement. Central authorities can use standard forms, treaty channels and agreed translations. Courts in the requested country can identify the applicable refusal grounds. The requested person can identify the precise allegations, the legal classification and the safeguards that the requesting State must satisfy. None of those benefits removes the judicial stage; they make that stage more structured.

That is also why an Indian treaty with a non-European State should not be described as a European arrest warrant. Within the European Union, the European arrest warrant is built on judicial cooperation and mutual recognition. A treaty with Kenya, Kazakhstan, Kyrgyzstan or Seychelles will remain an ordinary extradition instrument unless its text creates a different mechanism. There is no basis for assuming that a French court would replace the ordinary extradition chapter with the European arrest-warrant regime merely because India has expanded its treaty network. The legal route follows the relationship between the requesting and requested States, not the political intensity of the announcement.

For investigators, the practical consequence is a higher premium on preparation. A request that identifies the person but does not establish the identity reliably, gives an incomplete chronology, omits the applicable criminal provisions or fails to address the sentence sought will create avoidable delay. A treaty can reduce diplomatic friction, but it cannot repair a defective file. The request must also be compatible with the law where the person is found. In France, that means the prosecutor general, the chambre de l’instruction and, in the contested cases, the government will each have a defined role.

For the person sought, the first urgent question is not whether India has “a treaty” with the country of residence. It is which treaty, in which version, on which date, with which entry-into-force evidence and which request documents. The second question is whether the request is for prosecution or execution of a sentence. The third is whether the alleged conduct satisfies the treaty’s threshold and is an offence in both jurisdictions. A signature may start the clock of diplomatic work. It does not decide those questions.

B. Will a treaty secure surrender in economic-crime and terrorism cases?

The reported rationale for the new treaties includes terrorism, drug trafficking and economic crime. These categories are commercially and politically prominent, but they are not legal shortcuts. The requested State will still examine the exact conduct rather than accept the requesting State’s label. “Economic crime” can cover fraud, money laundering, tax offences, breach of trust or conduct regulated differently in each country. “Terrorism” can describe violence, financing, conspiracy or a political accusation. The treaty’s offence definitions, double-criminality clause, penalty threshold and exceptions will decide how a court treats the request.

The France–India convention demonstrates why labels are insufficient. Its Article 2 provides that extradition offences are facts punishable under the laws of both States by at least two years’ imprisonment. For a convicted person, the remaining prison term must be at least nine months. The convention also addresses tax, customs and exchange offences, and permits the requested State in certain multi-count cases to extend surrender to additional offences. The published language is precise: “Donnent lieu à extradition les faits punis par les lois des deux Etats contractants d’une peine d’emprisonnement d’au moins deux ans”. The question is therefore not simply whether an Indian warrant says “fraud”. It is whether the factual conduct, as described, meets the convention’s legal threshold in both systems.

That analysis matters particularly for financial cases. A request may rely on a complex series of transactions, a corporate structure, digital records and a charge that has no exact French equivalent. The court does not conduct the trial, but it must be able to map the conduct to the legal conditions for extradition. The requesting State should identify the acts, dates, locations, participants, victims and applicable provisions. If the evidence is framed only as a conclusion, the defence can argue that the court cannot test the treaty conditions or the corresponding offence under French law.

Political-offence safeguards remain relevant even when a request is described as a national-security matter. Article 3 of the France–India convention excludes extradition where the offence is considered political or connected to a political offence. It also permits the requested State to refuse where there are serious reasons to believe that the request seeks prosecution or punishment on grounds such as race, religion, nationality or political opinions, or that the person’s situation would be aggravated for one of those reasons. The convention carves out grave violence, but that does not turn every allegation made in a political conflict into an extraditable offence. The factual and legal characterisation remains necessary.

Nationality can also change the outcome. Article 5 of the France–India convention states that neither contracting State extradites its own nationals and measures nationality at the date of the facts. The requested State may instead be asked to submit the case to its own authorities for prosecution, subject to its domestic law and the requesting State’s request. The same issue appears in the French Code of Criminal Procedure. Article 696-4 places French nationality, assessed at the time of the offence, among the grounds against extradition. An Indian request directed at a person who later acquired French nationality therefore cannot be analysed by looking only at today’s passport.

Finality and time also matter. The France–India convention refuses extradition after a final judgment in the requested State for the same offence, where limitation has expired under either State’s law, and in certain amnesty situations. A request for an old economic offence must therefore be checked against both limitation systems and the procedural history in both countries. A treaty can supply an obligation to cooperate while preserving the principle that a person should not be transferred for a case already finally resolved or legally extinguished.

The figures reported in India also need careful reading. Hindustan Times reported that 25 fugitives had been successfully extradited to India in the previous five years, while 274 people had been brought back from 36 countries between 2019 and 2026, most through deportation. Those figures describe different legal routes. Extradition follows a formal request, arrest or detention measures, a judicial or administrative decision in the requested State and the treaty or statutory conditions governing surrender. Deportation may use immigration powers or coordinated removal without the same extradition litigation. Treating the two numbers as one measure of treaty performance would overstate what a new treaty has achieved.

It would also obscure the legal choices available to the requested State. A country may cooperate with India through deportation in one case while declining extradition in another because the treaty threshold is not met, the evidence is incomplete, the offence is political, the person is a national, or the proposed punishment conflicts with fundamental rights. Those outcomes are not contradictory. They show that the route matters.

India’s wider treaty network can nevertheless have a significant effect. It may deter fugitives from treating a jurisdiction as a safe location, give investigators a more stable framework for cross-border asset and evidence work, and make high-value cases easier to present. The likely pressure point will move from “is there any legal channel?” to “can the requesting State satisfy every condition in the channel?” In cases involving alleged corruption, cyber-enabled fraud, organised crime or politically sensitive conduct, that is where surrender will be won or lost.

The correct forecast is therefore conditional. A treaty makes a surrender request legally thinkable and diplomatically actionable. It does not compel a court to accept an incomplete file, does not eliminate human-rights review and does not make a deportation statistic proof of extradition success. For an accused person, the treaty is both a risk and a source of procedural leverage: the State has a stronger route to request transfer, but the route has rules that can be tested.

II. What would France examine before surrendering a person to India?

A. How do French courts test evidence, dual criminality and fundamental rights?

If the person sought is located in France, the starting point is the French extradition chapter, subject to the applicable bilateral convention. The version of the Code of Criminal Procedure displayed by Légifrance in August 2026 keeps the ordinary extradition provisions in force until the end of 2028, pending the scheduled legislative transition. The bilateral treaty comes first for matters it regulates; the Code supplies the French procedure and the rules that fill the gaps. This hierarchy is why the India–France convention and Articles 696 onward must be read together.

Readers looking for the wider French framework can also consult our French extradition procedure background. The present analysis narrows that framework to the legal consequences of India’s expanding treaty network and to the questions that a new India-related request would put before French authorities.

The threshold is set out in Article 696-3 of the Code of Criminal Procedure. It covers crimes under the requesting State’s law and, for misdemeanour-level offences, generally requires a maximum sentence of at least two years, or a sentence of at least two months where the person has already been convicted. The provision adds a French safeguard: extradition cannot be granted by the French government if the conduct is not punishable under French law by a criminal or correctional penalty. That is dual criminality in practical form. The names of the offences may differ; the conduct must still have a criminal counterpart and meet the applicable threshold.

The requested person should therefore compare the factual narrative with the two legal systems, not merely translate the charge. Was the conduct committed in the place stated? Does the French equivalent require an element missing from the request? Is the proposed sentence the maximum or the sentence actually imposed? Is the request for prosecution or for execution? Are several counts being aggregated to meet the threshold? Each detail can affect the court’s ability to give a lawful opinion.

The central refusal provision is Article 696-4. Its opening formula is categorical: “L’extradition n’est pas accordée”. The article addresses French nationality, political offences or a political purpose, offences committed in France, final proceedings in France, limitation, punishment contrary to French public policy, lack of fundamental procedural and defence guarantees, and military offences that are not ordinary-law offences. An Indian request must therefore be assessed against the full list rather than one broad question about diplomatic relations.

The evidence question is limited but real. French extradition proceedings are not a trial of guilt, and the French court does not decide whether the person committed the Indian offence. It does, however, need a coherent request from which it can test the statutory and treaty conditions. Article 696-8 requires a request through the diplomatic channel, supported by a judgment or conviction, a procedural act sending the person before a criminal court, or an arrest warrant or equivalent judicial act. The act must indicate precisely the facts and their date. The requesting government must also provide the relevant criminal-law texts. This is a document requirement, not a formality that can be ignored because the accusation is politically important.

After the documents are checked, Article 696-9 sends the file from the foreign affairs minister to the justice minister, who verifies the regularity of the request and refers it to the territorially competent procureur général. The route itself helps identify where a defence must act: at the diplomatic-document stage, at the arrest stage, before the prosecutor general, before the chambre de l’instruction and, if the government authorises extradition, before the competent administrative court.

French law also protects the person immediately after arrest. Under Article 696-10, a person arrested following an extradition request must be brought before the prosecutor general within forty-eight hours. The person must be told, in a language understood by that person, about the request and its content and may be assisted by chosen or appointed counsel. The prosecutor general must also explain the choice between consenting to and opposing extradition, and the consequences of waiving the rule of specialty. The article makes the record of the information a condition of regularity by stating that it must be entered in the report on pain of nullity.

Detention is not automatic either. Article 696-11 provides for presentation to the first president of the court of appeal or a designated judge where the prosecutor general does not leave the person at liberty. The judge can order extradition detention, but can also impose control measures where appearance at procedural acts is sufficiently guaranteed. A person who has a family, a fixed address, employment, medical needs or other evidence relevant to flight risk should present that material immediately; the question is not the merits of the Indian prosecution but the conditions under which the French procedure can continue.

In contested proceedings, the chamber of the instruction has a defined timetable and a defined object. Article 696-13 governs the appearance where consent is maintained. If the person does not consent, Article 696-15 requires the chamber to be seized without delay, provides for appearance within ten working days from presentation to the prosecutor general, and directs the chamber to give a reasoned opinion. The opinion is unfavourable if the legal conditions are not met or there is an obvious error. The text states: “la chambre de l’instruction donne son avis motivé sur la demande d’extradition”. A refusal is not an invitation to the government to disregard the court: Article 696-17 provides that a final unfavourable opinion prevents extradition.

Recent Cour de cassation decisions show how those statutory guarantees work in practice. In the case recorded under appeal no. 21-80.942, the Criminal Chamber held that the chamber of the instruction cannot give a favourable opinion without answering the necessity and proportionality of the measure in relation to the person’s private and family life. The Court’s formula was: “la chambre de l’instruction ne peut donner un avis favorable à l’extradition sans répondre sur la nécessité et la proportionnalité de la mesure au regard de l’atteinte portée au droit au respect de la vie privée et familiale de la personne réclamée”. The defence must therefore put the family and integration evidence before the court in a concrete way; a general assertion that the person has relatives in France will not do the work of a documented proportionality argument.

In appeal no. 20-82.296, the Court addressed conditions of detention and the information available from the requesting State. The decision criticised an opinion based only on a general assurance that detention would comply with international conventions, holding that the chamber should seek precise information where the file raises a concrete issue. The decision refers to the need for “informations précises auprès des autorités russes”. The principle is transferable to an India-related request: a defence based on a named prison, a medical condition, a documented risk or a sentence-specific problem should be answered with current and individualised material, not with a slogan about the requesting State’s general obligations.

These decisions do not create a merits trial in France. They insist that the court exercise the review assigned to it. A surrender opinion must be reasoned against the actual file, the actual person and the actual risk. That is the legal difference between a treaty-based request that is complete and one that relies on political confidence alone.

B. Why do death-penalty assurances, specialty and procedure still decide the case?

Even a complete request can fail on the destination sentence or the conditions of trial and detention. Article 696-4 treats punishment contrary to French public policy and the absence of fundamental procedural and defence guarantees as refusal grounds. Where the requesting State’s law permits capital punishment for the conduct alleged, the file must deal with that risk directly. A promise that is general, political or detached from the specific person may not be enough.

The leading modern illustration is the Criminal Chamber’s decision of 30 January 2024 in appeal no. 23-83.549. The Court held that, when the death penalty is available for the facts, the requesting State must give an undertaking “dénué de toute ambiguïté” that the punishment will not be applied to the specific person sought. The Court rejected reliance on general considerations such as a long moratorium and international commitments described as symbolic. The reasoning is especially important for a fast-moving extradition file: counsel should demand the exact assurance, the competent authority, the offence and the legal effect of the commitment. A statement that executions are rare is not necessarily the same as a binding undertaking for the individual case.

The France–India convention supplies its own treaty safeguards. Its Article 3 excludes political offences and requests driven by discriminatory or political considerations. Its Article 5 protects nationals. Its Article 6 addresses final judgments, limitation and amnesty. Its Article 7 lists situations in which extradition may be refused, including proceedings in the requested State or conduct partly committed there. Its Article 16 protects specialty: after delivery, the person may not be prosecuted, tried, detained to serve a sentence or otherwise deprived of liberty for an earlier offence other than the offence for which extradition was granted, subject to the treaty exceptions.

The same rule appears in Article 696-6 of the Code of Criminal Procedure, which makes extradition conditional on the person not being prosecuted or convicted for another earlier offence. Specialty is not technical wording reserved for diplomats. It defines the bargain on which surrender rests. If India requests transfer for one set of facts and then seeks to prosecute the person for a different earlier case, the requesting authorities must return to the treaty and obtain the required consent or satisfy an applicable exception. A defence should compare the warrant, the request, the conviction and any public statements about future prosecution.

Procedure continues after the judicial opinion. If the chamber gives a favourable opinion in an ordinary contested case, Article 696-18 provides that extradition is authorised by a decree of the Prime Minister, on the report of the justice minister, except where the statutory refusal mechanism applies. The same article sets a one-month period for a challenge for excess of power against the decree. The architecture is therefore dual: the judicial phase tests the legal conditions and gives an opinion; the executive phase decides whether to authorise surrender within the framework of the treaty and the court’s opinion; the administrative court can review the decree.

The time limits can matter as much as the merits. A provisional arrest is designed to hold the person while the formal file travels. The France–India convention provides that provisional arrest ends if the requested State has not received the extradition request and the supporting documents within sixty days, while the French Code contains its own rules for provisional arrest and release. A file that arrives late or without the required documents can therefore create an immediate release issue, even if the underlying allegation is serious.

The person’s nationality, residence and family situation should be documented at the beginning, not after an adverse opinion. The same is true of health evidence, detention conditions and the risk of political persecution. The decision in appeal no. 21-80.942 shows that the private and family-life argument must be proportionality-based. Appeal no. 20-82.296 shows that a detention argument should be tied to a concrete place and a request for specific information. Appeal no. 23-83.549 shows that a capital-punishment argument should identify the individual assurance that is missing or ambiguous. Those are litigation instructions that arise from the holdings, not from the headline attached to the request.

There is also a difference between the requested person’s challenge and an attempt to retry the foreign investigation in France. The chamber of the instruction does not decide whether an Indian court should convict. The defence must instead target the legal gateway: an unreliable identity, an incomplete warrant, a mismatch between facts and offences, an expired prosecution, a political purpose, a sentence contrary to French public policy, fundamental procedural guarantees, or a breach of treaty specialty. Precision makes the review effective. A broad declaration that the Indian prosecution is unfair, without evidence tied to the individual and the request, will be easier to dismiss than a documented challenge to a required condition.

For an Indian prosecutor, the answer is equally practical. The dossier should include a verified identity package; the judicial act or conviction; a factual chronology; the complete criminal provisions and penalty range; the status of any limitation period; the requested legal purpose; the treatment of nationals; the treatment of political conduct; the detention location; the trial guarantees; and, where relevant, a clear undertaking on capital punishment. It should also explain any parallel proceedings, the intended charge after transfer and the specialty position. The more the case touches financial crime, politics or national security, the less safe it is to rely on a generic statement of cooperation.

That is the broader significance of India’s new treaty strategy. A treaty can make cross-border enforcement more credible, but credibility in court comes from compliance with the instrument and the requested State’s law. In France, the route is visible: a formal request, document verification, arrest safeguards, counsel, the chambre de l’instruction, possible review of rights and proportionality, an executive decree and a challenge period. The path can be fast in urgent cases, but it is not automatic.

The immediate lesson for families and businesses is not to wait for the final surrender decision before obtaining the file. A red notice, diplomatic request, local arrest, immigration measure or press report can involve different legal mechanisms. The right response depends on identifying which mechanism has actually been triggered. A deportation order is not an extradition request. An extradition request is not a conviction. A treaty signature is not entry into force. A favourable opinion is not the same as physical delivery. Keeping these distinctions clear is often the difference between a theoretical argument and a procedural remedy available in time.

Conclusion

India’s reported approval of new extradition treaties with Kenya, Kazakhstan and the Kyrgyz Republic, together with the approval of the treaty signed with Seychelles, signals a serious attempt to enlarge the country’s formal reach over fugitives. It can improve the quality and predictability of future requests, especially in terrorism, drug-trafficking and economic-crime investigations. It can also reduce reliance on informal returns that are difficult to compare and harder to assess from a rights perspective.

But the legal effect of the announcement is conditional. The treaties must be signed, brought into force and used through their own conditions. The requesting State must produce a precise file. The requested State must apply its law. In France, dual criminality, nationality, limitation, political purpose, fundamental guarantees, family-life proportionality, detention conditions, death-penalty assurances and specialty can all change the result. The French courts do not decide the foreign guilt question, but they do control the gateway to surrender.

The headline is therefore best expressed as a warning: a signature creates a route, not a result. Anyone affected by an Indian extradition request should identify the treaty version, verify the procedural stage, obtain the underlying documents and challenge the exact legal condition that is missing. That is how the new network will be tested in practice—one file, one person and one requested State at a time.

📄 Official circular